This information is current as
of June 18, 2017.
Identification of Resting and Type I
IFN-Activated Human NK Cell miRNomes
Reveals MicroRNA-378 and MicroRNA-30e
as Negative Regulators of NK Cell
Cytotoxicity
Pin Wang, Yan Gu, Qian Zhang, Yanmei Han, Jin Hou, Li
Lin, Cong Wu, Yan Bao, Xiaoping Su, Minghong Jiang,
Qingqing Wang, Nan Li and Xuetao Cao
Supplementary
Material
References
Subscription
Permissions
Email Alerts
http://www.jimmunol.org/content/suppl/2012/05/30/jimmunol.120060
9.DC1
This article cites 33 articles, 15 of which you can access for free at:
http://www.jimmunol.org/content/189/1/211.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2012 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
J Immunol 2012; 189:211-221; Prepublished online 30 May
2012;
doi: 10.4049/jimmunol.1200609
http://www.jimmunol.org/content/189/1/211
The Journal of Immunology
Identification of Resting and Type I IFN-Activated Human
NK Cell miRNomes Reveals MicroRNA-378 and
MicroRNA-30e as Negative Regulators of NK Cell Cytotoxicity
Pin Wang,*,1 Yan Gu,*,1 Qian Zhang,*,1 Yanmei Han,* Jin Hou,* Li Lin,† Cong Wu,†
Yan Bao,* Xiaoping Su,* Minghong Jiang,‡ Qingqing Wang,† Nan Li,* and Xuetao Cao*,‡
M
icroRNAs (miRNAs, or miRs) exert functions mainly
through targeting the 39-untranslated region (39-UTR)
of mRNA to induce translation suppression or mRNA
degradation. It is well accepted that miRNAs participate in many
physiological and pathological processes (1), and the emerging
roles of miRNAs in immune development and response have been
extensively explored in recent years (2). Innate immunity, mainly
mediated by innate cells, including monocytes/macrophages, dendritic cells (DCs), granulocytes, and NK cells, provides the first line
of host defense against invading pathogens. It has been proven that
miRNAs play an indispensable role in innate immune responses
via regulating the development of monocytes/macrophages (3),
DCs (4), and granulocytes (5, 6), and also in regulating their responses to pathogens (7), TLR ligands (8), and viral infection (9).
Whereas NK cells are of vital importance in innate immune
*National Key Laboratory of Medical Immunology and Institute of Immunology,
Second Military Medical University, Shanghai 200433, China; †Institute of Immunology, Zhejiang University School of Medicine, Hangzhou 310058, China; and
‡
National Key Laboratory of Molecular Biology, Chinese Academy of Medical Sciences, Beijing 100730, China
1
P.W., Y.G., and Q.Z. contributed equally to this work.
Received for publication February 24, 2012. Accepted for publication May 2, 2012.
This work was supported in part by National High-Biotech Program of China Grants
2012AA020901 and 2012ZX10002014, National Natural Science Foundation of
China Grant 81123006, and National Key Basic Research Program of China Grant
2012CB910202.
Address correspondence and reprint requests to Dr. Xuetao Cao, National Key Laboratory of Medical Immunology and Institute of Immunology, Second Military Medical University, 800 Xiangyin Road, Shanghai 200433, China. E-mail address:
[email protected]
The online version of this article contains supplemental material.
Abbreviations used in this article: DC, dendritic cell; EGFP, enhanced GFP; GO,
gene ontology; miR, microRNA; miRNA, microRNA; RISC, RNA-induced silencing
complex; smRNA, small RNA; TPM, transcripts per million; 39-UTR, 39-untranslated region.
Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.1200609
responses against viral infection, contributing to the first line of
host antiviral defense, the function of intrinsic miRNAs in NK
cells remains elusive.
NK cells are a unique subset of innate immune cells that exhibit
cytotoxicity function such as adoptive CTL through a ligandreceptor recognition manner. NK cells mediate cytolysis through
a variety of mechanisms, including perforin/granzyme granulemediated exocytosis and activation of the TNF death receptor
family members. Studies of gene-disrupted mice and individuals
with inherited defects indicate that perforin and granzyme B are
vital to the cytotoxicity function of NK cells, and they play indispensable roles in granzyme-mediated cellular apoptosis. The
cytotoxicity of NK cells can be enhanced by many activating
cytokines, and type I IFN is one of the most potent activators of NK
cells both in humans and in mice (10), although the underlying
mechanisms remain to be fully understood. Type I IFN was shown
to stimulate accessory DCs to produce IL-15 and then to activate
NK cells (11), whereas type I IFN signaling-induced granzyme B
and perforin expression in murine NK cells was found to be required for NK cell activation in response to vaccinia virus infection (12). It seems that type I IFN activates NK cells in both direct
and indirect manners. It has been reported that type I IFN can
modulate miRNA expression in macrophages (13), and thus
whether type I IFN could regulate miRNA expression in NK cells
and subsequently regulate NK cell activity needs to be explored.
In-depth analysis of genome-wide miRNA expression profile
is a powerful approach to explore the function of miRNAs in a
given biological context. At present, miRNome is yielded mainly
through four kinds of high-throughput technology: the short RNA
cloning technique (14), the microarray technique (15), a new highthroughput microfluidic real-time quantitative PCR approach (16),
and the deep sequencing platform (17). Compared to the first three
high-throughput techniques, the second generation sequencing
has unparalleled advantages in fully exploring the small RNA
(smRNA) transcriptome including miRNome in a given tissue or
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
NK cells are important innate immune cells with potent cytotoxicity that can be activated by type I IFN from the host once infected.
How NK cell cytotoxicity is activated by type I IFN and then tightly regulated remain to be fully elucidated. MicroRNAs (miRNAs,
or miRs) are important regulators of innate immune response, but the full scale of miRNome in human NK cells remains to be
determined. In this study, we reported an in-depth analysis of miRNomes in resting and IFN-a–activated human NK cells, found
two abundant miRNAs, miR-378 and miR-30e, markedly decreased in activated NK cells by IFN-a, and further proved that miR378 and miR-30e directly targeted granzyme B and perforin, respectively. Thus, IFN-a activation suppresses miR-378 and miR30e expression to release cytolytic molecule mRNAs for their protein translation and then augments NK cell cytotoxicity.
Importantly, the phenomena are also confirmed in human NK cells activated by other cytokines and even in the sorted CD16+
CD56dimCD69+ human NK cell subset. Finally, miR-378 and miR-30e were proved to be suppressors of human NK cell cytotoxicity. Taken together, our results reveal that downregulated miR-378 and miR-30e during NK cell activation are negative
regulators of human NK cell cytotoxicity, providing a mechanistic explanation for regulation of NK cell function by
miRNAs. The Journal of Immunology, 2012, 189: 211–221.
212
miR-378 AND miR-30e CONTROL NK CELL CYTOTOXICITY
cell type under a certain physiological or pathological context.
Besides revealing the expression profile of known miRNAs, vast
amounts of sequencing data enable exploration of novel miRNA
candidates in a specific tissue or cell type and investigation of the
variation of mature miRNA sequences, which is called isomiRNA.
Also, it presents clues to investigate other kinds of smRNAs, for
instance, repeat-associated smRNAs and PIWI-interacting RNAs.
In order to investigate the roles of miRNAs in human NK cell
activation and function, we profiled smRNA components and
established miRNomes of resting or IFN-a-activated human periphery blood NK cells by deep sequencing. On the basis of comprehensive understanding of miRNAs profiling in human NK
cells, we found that miR-378 and miR-30e, two of the markedly
downregulated miRNAs in the activated human NK cells, directly
suppressed the expression of granzyme B and perforin and their
decreases during IFN-a activation granted human NK cells to be
more potent cytolytic, thus providing a new mechanistic explanation for the regulation of NK cell activation by miRNAs.
Cell culture and reagents
PBMCs were isolated from periphery blood of healthy donors using Ficoll
(Sigma-Aldrich, St. Louis, MO), and then the cells were labeled with
fluorescent Abs according to different experimental purpose. Finally, target
cellular subsets were sorted by MoFlo (Dako, Glostrup, Denmark). All
human blood samples were collected with the informed consent of healthy
donors and the experiments were approved by the Ethics Committee of the
Second Military Medical University, Shanghai, China. Fluorescent Abs were
all purchased from R&D Systems (Minneapolis, MN). Sorted human NK
cells (CD56+CD32) were cultured in IMDM medium (Life Technologies/
BRL, Carlsbad, CA) supplemented with 10% FBS (PAA Laboratories,
Pasching, Australia) and 10% human AB serum (TBDscience, Tianjin,
China). Human NK92 cell line, K562 cell line, and HeLa cell line were
obtained from the American Type Culture Collection (Manassas, VA) and
were cultured as described previously (18). Recombinant human IFN-a was
from R&D Systems and cycloheximide was from Sigma-Aldrich.
smRNA library generation and data production
RNA fractions enriched for smRNAs were isolated from cell pellets treated
with RNAlater (Life Technologies) using the mirVana miRNA isolation kit
(Life Technologies). smRNAs between ∼18 and 30 nt were isolated and
subjected to sequencing library production (Illumina, San Diego, CA) after
passing quality control. smRNA-Seq clean reads were mapped and classified according to annotations based on Rfam (19) miRBase (v.14.0) and
the National Center for Biotechnology Information GenBank noncoding
RNA database. The smRNAs that were not matched to any kind of categories were used to predict novel miRNAs.
Gene ontology term enrichment analysis of miRNA targets
Conserved miRNAs across the five species described in miRNA conservation analysis are equally divided into five ranks according to their expression level, with a higher rank number for a higher miRNA expression
level. Conserved targets, which are predicted by TargetScan (release 5.0)
Table I.
Prediction of novel miRNAs
The prediction of human novel miRNAs was implemented in the Mireap
program developed by the Beijing Genomics Institute. To identify atypical
and novel sequences of miRNAs in human, we adopted the following
strategy. First, candidate miRNA sites were screened out from breakpoints
defined by mapping of the smRNAs. Next, a minimal stringent criterion was
used to select miRNA candidates, which ensured that most sequences recovered were known miRNAs. Finally, the RNA secondary structure was
checked using Mfold.
Western blotting
Cells were lysed with M-PER protein extraction reagent (Pierce, Rockford,
IL) supplemented with protease inhibitor mixture, and protein concentrations of the extracts were measured by bicinchoninic acid assay (Pierce).
Forty micrograms of the protein was used for immunoprecipitation or was
loaded per lane, subjected to SDS-PAGE, transferred onto nitrocellulose
membranes, and then blotted (20).
Abs specific to human perforin and HRP-coupled secondary Abs were
obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Human
granzyme B Ab was obtained from Cell Signaling Technology (Beverly,
MA), and b-actin Ab was purchased from Bioworld Technology (St. Louis,
MO).
RNA quantification
Total RNA containing miRNA was extracted, reverse-transcribed, and
quantitative real-time PCR was amplified as described previously (9, 21).
For miRNA analysis, the reverse transcription primers for has-miR-30e
was 59-GTC GTA TCC AGT GCA GGG TCC GAG GTA TTC GCA
CTG GAT ACG ACG CTT CCA-39. Real-time PCR primers were 59-GGC
TGT AAA CAT CCT TGA C-39 (forward) and 59-GTG CAG GGT CCG
AGG T-39 (reverse). Reverse transcription primers and real-time PCR
primers for has-miR-378 were purchased from Ribobio (Guangzhou,
China). U6 small nuclear RNA was quantified using its reverse primer for
reverse transcription reaction and its forward and reverse primers for
quantitative PCR, which were 59-CTC GCT TCG GCA GCA CA-39
(forward) and 59-AAC GCT TCA CGA ATT TGC GT-39 (reverse). The
relative expression level of miRNAs was normalized to that of internal
control U6 by using the 22DDCt method and then multiplied with a certain
value. Primers for human granzyme B were 59-CCC TGG GAA AAC ACT
CAC ACA-39 (forward) and 59-CAC AAC TCA ATG GTA CTG TCG T-39
(reverse); primers for human perforin were 59-CCC AGT GGA CAC ACA
AAG GTT-39 (forward) and 59-TCG TTG CGG ATG CTA CGA G-39
(reverse); primers for human pri-miR-378 were 59-TGG AGT CAG AAG
GCC TGA GTG GAG T-39 (forward) and 59-GGA TTC TCT TGC GGC
ACC ACC A-39 (reverse); primers for human pri-miR-30e were 59-ACA
GCC ATT CTC CCC ATC CCA GC-39 (forward) and 59-CTG CCC AGA
AAG AAC GGA GAC G-39 (reverse); and primers for sponge mRNA were
59-TAC AAG TCC GGA CTC AGA TCT TAA-39 (forward) and 59-GGT
ACC GTC GAC TGC AGA ATT-39 (reverse). mRNA expression level was
normalized to the level of U6 expression in each sample.
Summary of data production by smRNA sequencing
NK-0 h
NK-12 h
NK-24 h
Type
Count
%
Count
%
Count
%
Total reads
High-quality readsa
Adaptor 39 null
Insert null
Adaptor 59 contaminants
Smaller than 18 nt
Poly(A)
Clean reads
16,862,552
12,470,391
99,590
246,533
177,270
483,023
60
11,463,915
100
0.80
1.98
1.42
3.87
0.00
91.93
17,982,973
14,582,582
297,216
436,647
276,972
631,276
56
12,940,415
100
2.04
2.99
1.90
4.33
0.00
88.74
17,826,283
13,777,178
95,287
260,607
291,897
780,246
92
12,349,049
100
0.69
1.89
2.12
5.66
0.00
89.63
a
In these reads, those containing ambiguous bases were removed from total reads.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Materials and Methods
(http://www.targetscan.org), of each rank are performed for enrichment
assay with hypergeometric test for gene ontology (GO):0002376. Variations for hypergeometric test are as follows: N, targets of analyzed total
miRNAs in upper GO term (GO:0008150, biological process); n, targets of
analyzed total miRNAs in target GO term (GO:0002376, immune system
response); M, targets of miRNAs of each rank in upper GO term (as
above); m, targets of miRNAs of each rank in target GO term (as above).
The Journal of Immunology
213
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 1. Analysis and functional annotation of miRNome of human NK cells. (A) Pie chart for percentages of total reads for categorized smRNAs in
resting NK cells (NK-0 h). More information about NK-12 h and NK-24 h is directed to Table II. (B) Pairwise comparison of known miRNA expression
profiles in the three libraries (NK-0 h, NK-12 h, and NK-24 h). Scatter plots show the log 10 value of TPM for each miRNA in the three libraries as
indicated; each plot represents one miRNA. (C) Conserved miRNAs expressed in resting NK cells are averagely divided into five ranks according to their
expression levels (the higher the rank number, the higher miRNA expression level). GO term enrichment analysis of predicted conserved targets is performed for each rank. The p value for immune system response gene set (GO: 0002376) is shown (see details in Materials and Methods). (D and E)
Clustering of dynamic expression patterns of markedly changed (.2-fold or ,0.5-fold, TPM . 100 at least in one library) known miRNAs (D) and novel
miRNA candidates (E) in human NK cells after IFN-a stimulation are shown in radar figures. Blank lines from the center separate six categories of dynamic
patterns as indicated. One red dot stands for one miRNA, and the distance to the center represents the significance of changes. One gray circle represents 1fold for miRNA variance. (F) Conserved targets of each miRNA category described in (D) are predicted by TargetScan (release 5.0) (http://www.targetscan.
org). Percentage of genes in NK cell-mediated cytotoxicity pathway (Kyoto Encyclopedia of Genes and Genomes database: hsa04650) covered by these
predicted targets for each category is shown. (G) Human primary NK cells were stimulated with or without 40 U/ml recombinant human IFN-a in the
presence or absence of cycloheximide (CHX; 10 mg/ml) for 8 h. Cytotoxicity of these cells is measured using flow cytometry with K562 cells as target cells.
The data are shown as means 6 SD (n = 4) of four independent experiments using NK cells from 4 donors. *p , 0.05.
214
miR-378 AND miR-30e CONTROL NK CELL CYTOTOXICITY
Table II. Percentage distributions of smRNAs among different
categories
RNA categories
Transfection of NK cells
Freshly isolated human NK cells were electrotransfected using an Amaxa
human NK cell Nucleofector kit (Lonza, Basel, Switzerland) and an NK92
cell line was electrotransfected using Amaxa cell line Nucleofector kit R
(Lonza, Basel), according to the manufacturer’s instructions.
NK-0 h NK-12 h NK-24 h
Known miRNAs
77.40
rRNA, tRNA, scRNA, snRNA, snoRNA 16.65
mRNAs (exons/introns)
3.54
Repeat-associated smRNAs
1.03
Unannotated smRNAs
0.76
piRNA
0.61
53.30
42.02
2.92
0.72
0.52
0.52
67.14
28.14
3.05
0.85
0.66
0.16
Establishment of stably transfected NK92 cells
piRNA, PIWI-interacting RNA; rRNA, ribosomal RNA; scRNA, small cytoplasmic RNA; snRNA, small nuclear RNA; snoRNA, small nucleolar RNA; tRNA,
transfer RNA.
To establish stably transfected NK92 cells, G418 was added (1000 mg/ml)
48 h after transfection and maintained at 800 mg/ml in the following 3 wk
for positive selection. For stably transfected cells, the expressions of EGFP
were confirmed by FASC and the expressions of sponge RNA were confirmed by quantitative real-time PCR (Fig. 6A). Stably transfected NK92
cells were subsequently cultured in medium containing 500 mg/ml G418.
RNA-binding protein immunoprecipitation
miRNA mimic or inhibitor
Luciferase reporter plasmid constructs
Full-length 39-UTR fragments of human GZMB and PRF1 were amplified
from cDNA of human peripheral blood NK cells and cloned into a pMIRReport construct (Ambion, Austin, TX). The mutant 39-UTRs carrying a
mutated sequence in the complementary site for the seed region of miR378 or miR-30e, generated using the fusion PCR, were also cloned into the
pMIR-Report construct.
NK92 cells were lysed using a complete RNA lysis buffer from a Magna
RNA-binding protein immunoprecipitation kit (Millipore, catalog no. 17701). RNA-binding protein immunoprecipitation experiments were performed according to the protocol provided with the kit. Human AGO2 Ab
was purchased from Cell Signaling Technology (Beverly, MA).
Intracellular staining and flow cytometry
Cells were stained using indicated Abs with the Cytofix/Cytoperm kit
(eBioscience) according to the manufacturer’s instructions, as described
previously (23). Flow cytometric analysis was performed on a FACS LSRII
with FACSDiva software (BD Biosciences).
Assay for NK cell cytotoxicity
NK cell cytotoxicity was analyzed using flow cytometry as previously
described (23).
Sponge plasmid construction
Statistical analysis
miRNA sponge plasmids were constructed on the basic structure of plasmid
pEGFP-C1. Briefly, we synthesized two DNA fragments with lengths of
∼250 bp containing seven sponge sites (for miR-30e sponge plasmid) or
six sponge sites (for miR-378 sponge plasmid) following upstream translation stop code TAA, which was flanked with two restriction sites, BglII
and EcoRI. After double digestion and ligation, these two DNA fragments
were inserted at the MCS site following the enhanced GFP (EGFP)
encoding sequence. All constructs were confirmed by DNA sequencing.
Statistical significance was determined by a paired two-tailed Student t test,
with a p value , 0.05 considered to be statistically significant. Additionally, two-tailed Pearson correlation coefficient analysis was performed
using SPSS software (version 17.0; SPSS, Chicago, IL).
Results
Assay of luciferase reporter gene expression
Analysis of miRNomes in resting and IFN-a–activated human
NK cells
HeLa cells were cotransfected with 80 ng indicated luciferase reporter
plasmid, 40 ng pRL-TK-Renilla-luciferase plasmid (Promega, Madison,
WI), and indicated RNAs (final concentration, 20 nM) using JetSI-ENDO
transfection reagents (PolyPlus Transfection, Illkirch, France). After 24 h,
normalized luciferase activities were obtained as previously discribed (22).
Alternatively, HeLa cells, which express certain level of endogenous
miR-378 and miR-30e, were cotransfected with 80 ng indicated luciferase
reporter plasmid, 40 ng pRL-TK-Renilla-luciferase plasmid (Promega),
and 80 ng indicated sponge plasmid using JetPEI transfection reagents
(PolyPlus Transfection). After 36 h, luciferase activities were measured
as above.
CD56+CD32 NK cells freshly isolated from peripheral blood of
healthy donors were used as resting human NK cells (NK-0 h), or
activated by recombinant human IFN-a (40 U/ml) for 12 h (NK-12
h) or 24 h (NK-24 h). RNA components of these human NK cells
were purified and subjected to smRNA Solexa sequencing. More
than 107 clean reads for each sample were obtained (Table I). The
mapping and categorizing of clean reads revealed that most
smRNA reads were miRNAs in NK cells (Fig. 1A, Table II). Unannotated smRNAs that were not classified to any of these cate-
Table III. Twelve most abundant miRNAs in human resting NK cells and their TPM in libraries of resting
NK cells and NK cells activated by IFN-a for 12 or 24 h
miRNA
NK-0 h
NK-12 h
NK-24 h
hsa-miR-21
hsa-let-7f
hsa-let-7a
hsa-miR-140-3p
hsa-miR-146b-5pa
hsa-let-7ga
hsa-miR-142-3p
hsa-miR-101
hsa-miR-378a
hsa-let-7b
hsa-miR-103
hsa-miR-30ea
102,646.9
80,473.1
78,552.1
44,283.7
42,726.2
31,603.2
29,283.6
21,942.3
17,246.0
12,876.2
11,703.6
9,352.9
60,147.2
65,097.3
64,943.5
23,931.8
38,954.3
26,704.2
21,463.1
12,855.1
7,228.5
17,830.8
5,127.6
5,457.2
97,071.6
68,550.5
73,008.5
44,594.4
33,435.5
25,027.5
26,566.1
17,532.8
4,408.8
20,436.9
8,631.0
4,602.9
a
Continuously downregulated miRNAs during the process of IFN-a activation.
Maximum Variation Fold
0.59
0.81
0.83
0.54
0.78
0.79
0.73
0.59
0.26
1.59
0.44
0.49
(NK-12
(NK-12
(NK-12
(NK-12
(NK-24
(NK-24
(NK-12
(NK-12
(NK-24
(NK-24
(NK-12
(NK-24
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h/NK-0
h)
h)
h)
h)
h)
h)
h)
h)
h)
h)
h)
h)
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
miR-378 and miR-30e mimics and inhibitors from Ribobio were used for
the overexpression and inhibition, respectively, of miR-378 and miR-30e in
HeLa cells. Negative control mimic or inhibitor (Ribobio) was transfected
as the matched controls. miRNA mimic with Cy3 labeled at the end of its
antisense strand (Ribobio) was used for the overexpression of miR-378 and
miR-30e in human NK cells.
The Journal of Immunology
gories were used for novel miRNA candidate prediction. We
identified .200 novel miRNA candidates.
The abundance value of each known miRNA was normalized
using transcripts per million (TPM) in each smRNA library. Known
miRNAs expressed in these three libraries exhibited a similar
expression profile (Fig. 1B). Only ∼1.05% of registered miRNAs
were expressed abundantly (.10,000 TPM) but they accounted
for ∼76.8% of all miRNA reads. The 12 most abundant miRNAs
in the NK-0 h library are listed in Table III. miRNAs with more
abundance may be more related to the biological characteristics of
215
NK cells. Through GO enrichment analysis of predicted targets,
we found that predicted conserved targets of miRNAs with higher
expression levels were more enriched in the gene set of the immune system response term classified in the Gene Ontology database, indicated by the rising broken line in Fig. 1C. This
implied, at least partially, that immune cells with specific functions might highly express functionally related miRNAs, which
could be implicated in precise regulation of cell-specific gene
expression pattern and thus contribute to the identity of these cells
with specific functions.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 2. IFN-a regulates the expression of miR-378, miR-30e, and cytolytic molecules in human NK cells. Human peripheral blood NK cells (CD56+
CD32) were sorted from 15 healthy donors and then stimulated with recombinant human IFN-a (40 U/ml) for the indicated times. miR-378 and miR-30e
levels were detected using quantitative real-time PCR (A), and intracellular granzyme B and perforin expression was detected using flow cytometry (D).
The filled bar indicates the mean value of each treatment, and p , 0.01 for 0 versus 12 h and 0 versus 24 h. In those samples that exhibited a significant
decrease in (A) (,0.5-fold), NK cells were stimulated with recombinant human IFN-a at different concentrations as indicated for 24 h, and then the
expression of miR-378 and miR-30e was measured (B), and pri-miR-378 and pri-miR-30e levels were also detected by quantitative real-time PCR (C). In
samples exhibiting significant increase in (D) (.2-fold), granzyme B and perforin levels were detected by Western blot, with b-actin as an input control (E),
and their mRNA levels were examined by quantitative real-time PCR (F). Data are representative of five independent donors (B, C, E, F). We applied
a decrement value of miR-378 relative expression (subtract miR-378 relative expression value at 24 h from its value at 0 h) versus granzyme B mean
fluorescence intensity (MFI) increment value (subtract granzyme B MFI value at 0 h from the value at 24 h) (left) and decrement value of miR-30e relative
expression versus perforin MFI increment value (right) for correlation statistical analysis using the Pearson correlation coefficient (G). Data are from three
independent experiments (mean 6 SEM). *p , 0.05, **p , 0.01, ***p , 0.001.
216
miR-378 AND miR-30e CONTROL NK CELL CYTOTOXICITY
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 3. miR-378 and miR-30e directly target granzyme B and perforin, respectively. (A) Human GZMB and PRF1 are potential targets of miR-378
and miR-30e, respectively. Shown is a sequence alignment of miRNAs and their target sites in 39-UTRs of their respective targets and the mutant sites in the
following gene-reporting experiments. (B and C) HeLa cells (1 3 104) were cotransfected with 80 ng wild-type (WT) or mutant 39-UTR firefly luciferase
reporter plasmids, 40 ng pTK-Renilla-luciferase plasmids, together with miR-378 mimic or its control mimic (B, left), miR-378 inhibitor or its control
inhibitor (B, right), miR-30e mimic or its control mimic (C, left), or miR-30e inhibitor or its control inhibitor (C, right) at the final concentration of 20 nM.
After 24 h, firefly luciferase activity was measured and normalized by Renilla luciferase activity. (D) HeLa cells (1 3 104) were cotransfected with 80 ng
WT or mutant 39-UTR firefly luciferase reporter plasmids, 40 ng pTK-Renilla-luciferase plasmids, together with 80 ng miR-378 sponge plasmids or control
plasmids (left) or miR-30e sponge plasmids or control plasmids (right). After 36 h, firefly luciferase activity was measured and normalized by Renilla
luciferase activity. (E and F) HeLa cells (1 3 104) were cotransfected with 80 ng WT human GZMB 39-UTR firefly luciferase reporter plasmids (E) or 80 ng
WT human PRF1 39-UTR firefly luciferase reporter plasmids (F), 40 ng pTK-Renilla-luciferase plasmids, together with miRNA mimic as indicated at the
final concentration of 20 nM. After 24 h, firefly luciferase activity was measured and normalized by Renilla luciferase activity. (G and H) Freshly sorted
primary human NK cells (3 3 106) were eletrotransfected with miR-378 mimic, miR-30e mimic, or control mimic, all labeled with Cy3 (200 nM) (G) or
with 2 mg miR-378 sponge plasmid, miR-30e sponge plasmid, or control plasmid (H). Ten hours after transfection, cells were stimulated with or without
recombinant human IFN-a (40 U/ml) for 24 h. Then, Cy3-positive cells (PE-positive) and EGFP-positive cells (FITC-positive) were sorted for the detection
of cytolytic molecules and miRNA. The mean fluorescence intensity (MFI) value was quantified in line charts as well. Flow cytometry data are representative of three independent experiments, with each experiment using NK cells from one donor. (I) NK92 cells were cultured without IL-2 for 24 h and
then stimulated with recombinant human IFN-a (40 U/ml) for indicated time. RISC was immunoprecipitated with anti-human Ago2 Ab. Granzyme B and
perforin mRNA levels in this complex were normalized to that in 1% input control of their respective sample and then compared with that of the 0 h sample.
Data are shown as means 6 SD (n = 4) of one representative experiment. Similar results were obtained in at least three independent experiments. *p , 0.05.
**p , 0.01.
The Journal of Immunology
granzyme B were also observed in human NK cell line NK92 after
IFN-a activation (Supplemental Fig. 1A–C).
miR-378 and miR-30e directly target the 39-UTR of human
GZMB and PRF1, respectively
As predicted by computational prediction via TargetScan (http://
www.targetscan.org), human GZMB and PRF1 mRNAs are potential targets of miR-378 and miR-30e, respectively (Fig. 3A). To
verify these predictions, we constructed luciferase reporter plasmid with a 39-UTR sequence of human GZMB or PRF1, or with
their mutant 39-UTR sequence. Overexpression and inhibition experiments using synthetic miRNA mimics or inhibitors revealed
that miR-378 and miR-30e targeted the 39-UTR of GZMB and
PRF1, respectively, precisely at the indicated sites (Fig. 3B, 3C).
We applied another strategy of miRNA knockdown, that is,
“miRNA sponge” plasmids that express EGFP mRNA with many
bulged binding sites in its 39-UTR to absorb intrinsic miR378 or
miR30e (24). The miR-378 and miR-30e sponges enhanced
GZMB and PRF1 39-UTR luciferase activity, respectively, and the
enhancement was lost after site mutations (Fig. 3D).
miR-378 and miR-30e expression are negatively correlated
with granzyme B and perforin protein levels, respectively,
during IFN-a activation in human NK cells
Detection of the kinetic expression of miR-378 and miR-30e in
IFN-a–activated human NK cells from as many as 15 donors
showed varying degrees of decline (Fig. 2A). Using samples that
exhibited a significant decrease (,0.5-fold), we confirmed that
their decreases were IFN-a dose-dependent (Fig. 2B), and this
study also revealed that the expression of the primary transcripts
of miR-378 and miR-30e also decreased in IFN-a–activated NK
cells (Fig. 2C).
We found that the protein levels of granzyme B and perforin,
which are the predicted targets of miR-378 and miR-30e, respectively, increased in varying degrees after IFN-a activation by flow
cytometry assay (Fig. 2D). We chose the samples with significant
increases of granzyme B and perforin (.2-fold), which were confirmed by immunoblots (Fig. 2E), examined their mRNA levels,
and found no significant differences (Fig. 2F), indicating that there
is a posttranscriptional regulatory mechanism for IFN-a–upregulated granzyme B and perforin expression in human NK cells.
To further explore whether these two decreased miRNAs have
certain relevance to the increased cytolytic molecules, we utilized
IFN-a reactivity variation of samples from 15 donors (Fig. 2A,
2D) and performed correlation analysis for the value of miRNA
decreased levels versus the value of cytolytic molecule increased
levels of all human donors during IFN-a activation, and we found
a strong correlation (p , 0.001) (Fig. 2G). This result strongly
suggests the relevance between miRNA decrease and cytolytic
molecule increase.
Additionally, the decreasing expression patterns of miR-378 and
miR-30e versus increasing expression patterns of perforin and
FIGURE 4. The expression of miR-378 and miR-30e negatively correlates with expression of cytolytic molecule granzyme B and perforin in
human NK cells in vivo. (A) Freshly sorted human NK cells (CD56+CD32)
of 10 healthy donors were separately subjected to quantitative PCR assay
of miR-378 and miR-30e expression and flow cytometry assay of intracellular granzyme B and perforin expression. We used the DCt value of
miR-378 (subtract miRNA Ct value from U6 Ct value) and mean fluorescence intensity (MFI) value of granzyme B (left) and the DCt value of
miR-30e and the MFI value of perforin (right) for correlation statistical
analysis using the Pearson correlation coefficient. (B and C) CD8+CD3+
cells, CD56+CD3+ cells, CD162CD56hi cells, and CD16+CD56dim cells
were sorted from PBMCs of health donors. (B) The expression level of
miR-378 was detected as in Fig. 2A, and the level of granzyme B was
examined by flow cytometry for each of these cell subsets. (C) miR-30e
level and intracellular perforin were also detected as in (B). Data are
representative of five independent donors.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Next, we analyzed the dynamic expression patterns of markedly
changed miRNAs (1-fold increment or decrement, and TPM .100)
in NK cell miRNome upon IFN-a stimulation. Interestingly, we
found most of them were significantly downregulated in IFN-a–
activated NK cells, as indicated by intensive red dots in the lower
left area of the radar (Fig. 1D), although some of them increased
again when IFN-a stimulation extended to 24 h. However, few
miRNAs of NK cells increased during the activation. Additionally,
novel miRNA candidates (TPM . 100) showed little change
during IFN-a activation (Fig. 1E).
We further explored the biological significance of the changed
miRNAs during NK cell activation, especially in NK cell cytotoxicity. We subjected conserved targets of six groups of significantly changed miRNAs with different expression patterns (Fig.
1D) to the NK cell cytotoxicity pathway from the Kyoto Encyclopedia of Genes and Genomes database. We found that targets
of continuously downregulated miRNAs (category no. 6) were
most enriched in this pathway compared with other differentially
regulated miRNAs (Fig. 1F). This result implied that in IFN-a–
activated human NK cells, some miRNAs were downregulated
continuously to release mRNAs for protein translation. We also
found that blockade of protein translation by cycloheximide attenuated cytolytic activity induced by IFN-a (Fig. 1G) whereas
blockade of transcription by actinomycin D had little effect
(Supplemental Fig. 2), indicating that synthesis of new proteins
rather than transcriptional activation was necessary for the enhancement of human NK cell cytotoxicity after IFN-a activation.
Therefore, we wondered whether these continuously downregulated miRNAs could regulate NK cell cytotoxicity. Among them
we found that miR-378 and miR-30e, which were also among the
12 most abundant miRNAs in NK cells (Table III), were involved
in the regulation of NK cell cytotoxicity.
217
218
Besides miR-378 and miR-30e, we also investigated other
miRNAs predicted to target human GZMB or PRF1. Most of them
were seldomly expressed in human NK cells (TPM , 10) or
showed no apparent expression changes during IFN-a stimulation.
We chose 11 miRNAs that have certain abundance in human NK
cells (TPM . 100) to perform the luciferase reporter gene assay,
which showed that miR-378 and miR-30e were the most effective
regulators (Fig. 3E, 3F). The above data show that miR-378 and
miR-30e are the key regulatory miRNAs targeting GZMB and
PRF1, respectively, in human NK cells.
miR-378 and miR-30e suppress IFN-a–upregulated granzyme
B and perforin expression in human NK cells
FIGURE 5. miR-378 and miR-30e are
downregulated in human NK cells activated by IL-15, IL-12, and IL-2 in vitro
and CD16+CD56dimCD69+ human NK
cells in vivo. Human NK cells from five
donors were stimulated with IL-15 (100
ng/ml), IL-12 (40 ng/ml), a high dose of
IL-2 (1000 U/ml), or IL-2 (100 U/ml)
plus PHA (10 ng/ml) as indicated for
24 h, and then miR-378 and miR-30e expression (A) was detected as in Fig. 2A.
Also, intracellular granzyme B and perforin expression was detected by flow
cytometry (B) and Western blot (C) as in
Fig. 2E. CD69+ and CD692 subsets of
CD16+CD56dim NK cells in PBMCs
were sorted, their cytolytic molecules
were assayed by flow cytometry, and
miR-378/miR-30e expression level was
examined by real-time PCR. Data are
representative of four independent donors (mean 6 SEM) (C, D). *p , 0.05,
**p , 0.01.
cytolytic molecules might disassociate with the RNA-induced silencing complex (RISC) in NK cells after IFN-a stimulation. To
verify this, we immunoprecipitated RISC from human NK cell
line NK92 using anti-human AGO2 Ab before and after IFN-a
stimulation and then examined the level of cytolytic molecule
mRNAs. As expected, both granzyme B and perforin mRNAs
decreased in RISC after IFN-a stimulation (Fig. 3I). These data
demonstrated that miR-378 and miR-30e suppress IFN-a–upregulated granzyme B and perforin expression in human NK cells.
miR-378 and miR-30e participate in the regulation of
granzyme B and perforin expression in human NK cells in vivo
Given the variation among humans, we found that the expression
level of cytolytic molecules in freshly isolated human NK cells
varied a lot among individuals in our experiments. Is this variation
due to the different expression levels of endogenous miR-378 and
miR-30e in NK cells from different individuals? Correlation
analysis revealed that there was a negative relationship between
the miR-378 level and the granzyme B protein expression level in
peripheral blood NK cells among healthy donors (Fig. 4A, left).
The same negative relationship exists between miR-30e expression and the perforin protein expression level (Fig. 4A, right).
Furthermore, among different subsets of cytotoxic cells in one
individual, negative relationships were also found between cytolytic molecules and these miRNAs. As shown in Fig. 4B and 4C,
cytolytic NK cells (CD16+CD56dim) with higher effector molecule
expression had lower miR-378 and miR-30e expression, whereas
CD162CD56hi NK cells and NKT cells (CD56+CD3+) with a
lower effector molecule expression had higher miR-378 and miR30e expression.
These results imply that miR-378 and miR-30e control the
expression level of cytolytic molecules both among different
individuals and in different cytotoxic NK/NKT/CD8+ T cell subsets of one human body, indicating that this is a universal regulation mechanism of cytolytic molecule (granzyme B, perforin)
expression in vivo.
Universal downregulation of miR-378 and miR-30e in the
activated human NK cells
We wondered whether the enhancement of cytolytic molecule expression by decreasing miR-378 and miR-30e was specific for
type I IFN-mediated NK cell activation or a universal phenom-
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Next, we investigated whether miR-378 and miR-30e could suppress intrinsic expression of cytolytic molecules in human NK
cells. To overcome the low transfection efficiency problem in
human primary NK cells, we electrotransfected human primary NK
cells with miRNA mimics labeled with Cy3, a small fluorescent
molecule, at the end of the antisense strand, and selected the PEpositive cells (successfully transfected with Cy3-labeled mimic)
after IFN-a stimulation to analyze the expression of the target
molecules. Flow cytometry assays show that the miR-378 mimic
suppressed granzyme B expression and that the miR-30e mimic
suppressed perforin expression in primary human NK cells,
largely overcoming the effect of IFN-a on granzyme B and perforin, respectively (Fig. 3G, histograms and line charts), indicating
that miR-378 and miR-30e play a major role in IFN-a–induced
cytolytic molecule expression. The miRNA sponge plasmids
mentioned above were also used for loss-of-function experiments.
We took EGFP/FITC-positive cells for the expression detections
of target molecules and found that the miR-378 sponge increased
granzyme B expression and the miR-30e sponge increased perforin expression in human primary NK cells (Fig. 3H). In the line
chart, miRNA sponges made their specific targets increase more
sharply with IFN-a treatment, indicating that miRNA sponges
synergize with IFN-a on their specific target molecules.
Based on these results, we proved that miR-378 and miR-30e
could suppress mRNA translation of cytolytic molecules by directly targeting their 39-UTRs, and IFN-a stimulation might reverse the suppressive effect by downregulating expression of miR378 and miR-30e so as to release granzyme B and perforin
mRNAs from translational inhibition. In this manner, mRNAs of
miR-378 AND miR-30e CONTROL NK CELL CYTOTOXICITY
The Journal of Immunology
miR-378 and miR-30e inhibit cytotoxicity of human NK cells
Finally, to confirm the roles of miR-378 and miR-30e in the
negative regulation of NK cell cytotoxicity, we stably transfected
human NK cell line NK92 with miR-378 or miR-30e sponge
plasmid and confirmed the expressions of EGFP protein and sponge
mRNA (Fig. 6A). Increments of granzyme B in the miR378
sponge cell line and increments of perforin in the miR-30e sponge
cell line were observed (Fig. 6B). Accordingly, the NK92 cells
stably transfected with miR-378 sponge plasmid or miR-30e
sponge plasmid exhibited more potent cytolytic activity compared with NK92 cells stably transfected with control plasmid
(Fig. 6C). Therefore, we demonstrated that miR-378 and miR-30e
significantly suppress cytotoxicity of human NK cells by targeting
cytolytic molecules granzyme B and perforin.
Discussion
In-depth miRNA profiling by deep sequencing with high-resolution views of expressed miRNAs over a wide dynamic range of
expression levels has already been realized in several kinds of
tissues and cell types, including human embryonic stem cells (25),
human B cells (26), cell types during mouse lymphopoiesis (17),
human renal cell carcinoma (27), and mouse leukemia (28). The
study that revealed dynamic miRNomes during lymphopoiesis
selected mouse adherent lymphokine-activated NK cells (ALAK,
NK cells) as a mouse NK cell substitute to get the mouse NK cell
miRNome (17), which may affect the biological and physiological
significance of this mouse NK cell miRNome. In our study, by
deep sequencing, we performed the smRNA expression profiling
of the freshly sorted human peripheral blood NK cells during the
process of type I IFNs activation, paving the way to explore the
relationship between smRNAs, including miRNAs, and NK cell
biological function, which is of physiological relevance to deepen
our understanding of the mechanism of human NK cell activity.
FIGURE 6. Suppression of miR-378 and miR-30e enhances cytolytic
activity of human NK cells. Sponge-linked EGFP expression (A, left) and
sponge mRNA expression (A, right) was confirmed in the stably transfected NK92 cell lines by flow cytometry assay and real-time PCR, respectively. Shown is the flow cytometry plot of miR-378 sponge stable cell
line as representative. Intracellular granzyme B and perforin protein levels
in the miRNA sponge stable NK92 cell lines were detected by flow
cytometry (B). The cytolytic assay by flow cytometry with K562 cells as
target cells was done using cells transfected with single sponges or two
sponges together (C). Flow cytometry plots are representative of results of
three independent experiments. Data are shown as means 6 SD (n = 3) of
one representative experiment. Similar results were obtained in at least
three independent experiments. *p , 0.05.
Based on analysis of miRNomes, we went further to find that miR378 and miR-30e, abundantly expressed in human NK cells,
suppress NK cell cytolytic activity through targeting cytolytic
molecules, granzyme B and perforin, and the decreases of miR378 and miR-30e, a common phenomenon during human NK cell
activation, contribute to full activation of NK cells.
As we showed above, after IFN-a activation many miRNAs
decreased whereas few miRNAs increased, and the amount of
total miRNAs also decreased (Table II). It is reasonable that type I
IFN signaling downregulated miRNA expression in NK cells to
assist NK cells in the enhancement of some protein expressions to
augment NK cell cytolytic activity. It is probable that, apart from
miR-378 and miR-30e, we are far from discovering the last
miRNA taking part in the regulation of NK cell function, and we
cannot exclude the possibility that changed miRNAs in IFN-a–
activated NK cell feedback regulate the type I IFN signaling
pathway, which is vital to many immune responses, including NK
cell activation. To comprehensively assess the roles of miRNAs
in type I IFN signaling and NK cells activation, integration of
miRNome data with proteomics data should be helpful.
Compared with their counterparts in mice, we found that human
NK cells are less sensitive to activating cytokine stimulation and
have a higher basic expression level of cytolytic molecules. This is
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
enon for NK cell activation. We used cytokines IL-15, IL-12, and
IL-2 to activate human NK cells and then analyzed the expression
of miRNAs and cytolytic molecules. As shown in Fig. 5A, the
expression of miR-378 and miR-30e decreased to various degrees;
additionally, flow cytometry assays (Fig. 5B) and immunoblots
(Fig. 5C) both confirmed that the expression of cytolytic molecules was enhanced by all of the cytokines. Interestingly, IL-15
stimulation enhanced granzyme B and perforin expression most
markedly in NK cells, most significantly decreasing miR-378 and
miR-30e.
To confirm that the decrease of miR-378 and miR-30e also exists
in the physical process of human NK cell activation in vivo, we
freshly isolated the activated human NK cell subset (CD16+
CD56dimCD32 CD69+) and resting NK cells (CD16+CD56dim
CD32CD692) from the same donor sample to detect the expression level of miR-378 and miR-30e. Indeed, the expression of
these two miRNAs was lower in the activated NK cell subset
compared with that in resting NK cells from the same donor (Fig.
5D). Flow cytometry assays showed that CD69+ NK cells had a
higher level of perforin and a little higher leveler of granzyme B as
compared with CD692 NK cells (Fig. 5D). The data confirm that
the negative relationship between miR378/miR-30e and cytolytic
molecules also exists in the physical process of human NK cell
activation in vivo, together with other data further suggesting that
miR-378 and miR-30e negatively regulate granzyme B and perforin expression in human NK cells in vivo.
The finding that human NK cells downregulate miR-378 and
miR-30e expression to increase cytolytic molecule expression
shows a fundamental mechanism for the activation of human NK
cells, which is not confined to IFN-a-activation.
219
220
the identification of miRNA transcriptome in mouse splenic NK
cells (NK1.1+CD32) using two different kinds of second generation techniques. After thorough analysis of miRNome in resting
and IL-15–activited mouse NK cells, they found that miR-223
decreased following IL-15 activation and specifically targeted
the 39-UTR of murine GZMB in vitro. With regard to human NK
cells, we found that miR-223 was expressed at a very low level
(TPM , 30 in our human NK cell miRNome; data not shown) and
had no predicted targeting site in human GZMB 39-UTR, implying
that granzyme B expression was definitely regulated by miRNAs
in mammalian NK cells whereas human and mouse NK cells
applied different miRNAs in this regulation. Taken together, these
two works provide a comprehensive understanding of miRNome
in NK cells across mammalian species.
Acknowledgments
We thank Dr. Dong Li, Dr. Feng Ma, Jianqiu Long, and Xiaoting Zuo of the
National Key Laboratory of Medical Immunology and Institute of Immunology, as well as Fei Teng, Dr. Ning Li, and Dr. Jun Wang of Beijing
Genomics Institute–Shenzhen, for technical assistance.
Disclosures
The authors have no financial conflicts of interest.
References
1. Bushati, N., and S. M. Cohen. 2007. MicroRNA functions. Annu. Rev. Cell Dev.
Biol. 23: 175–205.
2. O’Connell, R. M., D. S. Rao, A. A. Chaudhuri, and D. Baltimore. 2010. Physiological and pathological roles for microRNAs in the immune system. Nat. Rev.
Immunol. 10: 111–122.
3. Fontana, L., E. Pelosi, P. Greco, S. Racanicchi, U. Testa, F. Liuzzi, C. M. Croce,
E. Brunetti, F. Grignani, and C. Peschle. 2007. MicroRNAs 17-5p-20a-106a
control monocytopoiesis through AML1 targeting and M-CSF receptor upregulation. Nat. Cell Biol. 9: 775–787.
4. Rodriguez, A., E. Vigorito, S. Clare, M. V. Warren, P. Couttet, D. R. Soond,
S. van Dongen, R. J. Grocock, P. P. Das, E. A. Miska, et al. 2007. Requirement of
bic/microRNA-155 for normal immune function. Science 316: 608–611.
5. Velu, C. S., A. M. Baktula, and H. L. Grimes. 2009. Gfi1 regulates miR-21 and
miR-196b to control myelopoiesis. Blood 113: 4720–4728.
6. O’Connell, R. M., A. A. Chaudhuri, D. S. Rao, and D. Baltimore. 2009. Inositol
phosphatase SHIP1 is a primary target of miR-155. Proc. Natl. Acad. Sci. USA
106: 7113–7118.
7. Johnnidis, J. B., M. H. Harris, R. T. Wheeler, S. Stehling-Sun, M. H. Lam,
O. Kirak, T. R. Brummelkamp, M. D. Fleming, and F. D. Camargo. 2008.
Regulation of progenitor cell proliferation and granulocyte function by microRNA-223. Nature 451: 1125–1129.
8. Sheedy, F. J., E. Palsson-McDermott, E. J. Hennessy, C. Martin, J. J. O’Leary,
Q. Ruan, D. S. Johnson, Y. Chen, and L. A. O’Neill. 2010. Negative regulation of
TLR4 via targeting of the proinflammatory tumor suppressor PDCD4 by the
microRNA miR-21. Nat. Immunol. 11: 141–147.
9. Wang, P., J. Hou, L. Lin, C. Wang, X. Liu, D. Li, F. Ma, Z. Wang, and X. Cao.
2010. Inducible microRNA-155 feedback promotes type I IFN signaling in antiviral innate immunity by targeting suppressor of cytokine signaling 1. J.
Immunol. 185: 6226–6233.
10. Biron, C. A., K. B. Nguyen, G. C. Pien, L. P. Cousens, and T. P. Salazar-Mather.
1999. Natural killer cells in antiviral defense: function and regulation by innate
cytokines. Annu. Rev. Immunol. 17: 189–220.
11. Lucas, M., W. Schachterle, K. Oberle, P. Aichele, and A. Diefenbach. 2007.
Dendritic cells prime natural killer cells by trans-presenting interleukin 15.
Immunity 26: 503–517.
12. Martinez, J., X. Huang, and Y. Yang. 2008. Direct action of type I IFN on NK
cells is required for their activation in response to vaccinia viral infection in vivo.
J. Immunol. 180: 1592–1597.
13. O’Connell, R. M., K. D. Taganov, M. P. Boldin, G. Cheng, and D. Baltimore.
2007. MicroRNA-155 is induced during the macrophage inflammatory response.
Proc. Natl. Acad. Sci. USA 104: 1604–1609.
14. Basso, K., P. Sumazin, P. Morozov, C. Schneider, R. L. Maute, Y. Kitagawa,
J. Mandelbaum, J. Haddad, Jr., C. Z. Chen, A. Califano, and R. Dalla-Favera.
2009. Identification of the human mature B cell miRNome. Immunity 30: 744–752.
15. Ceppi, M., P. M. Pereira, I. Dunand-Sauthier, E. Barras, W. Reith, M. A. Santos,
and P. Pierre. 2009. MicroRNA-155 modulates the interleukin-1 signaling
pathway in activated human monocyte-derived dendritic cells. Proc. Natl. Acad.
Sci. USA 106: 2735–2740.
16. Petriv, O. I., F. Kuchenbauer, A. D. Delaney, V. Lecault, A. White, D. Kent,
L. Marmolejo, M. Heuser, T. Berg, M. Copley, et al. 2010. Comprehensive
microRNA expression profiling of the hematopoietic hierarchy. Proc. Natl. Acad.
Sci. USA 107: 15443–15448.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
reasonable, because human NK cells encounter so many challenges
and stimulation from outside pathogens and the induced inflammatory responses, whereas specific pathogen-free laboratory
mouse NK cells seldom suffer such challenge. Thus, NK cells from
specific pathogen-free mice are globally naive, whereas activation
levels of NK cells from different humans are of high variation.
Different expression levels of miR-378 and miR-30e in NK cells
from different donors probably imply their different activating
level, which is supported by the accordingly different expression
levels of granzyme B and perforin (Fig. 4A). NK cells from some
donors with high cytolytic molecule expression and low miR-378
and miR-30e expression did not respond obviously to IFN-a
stimulation (Fig. 2A, 2D), probably because they have already
been activated by some challenge without apparent symptom.
Previous research showed that cytotoxic activity of the CD162
CD56hi human NK cell subset increases significantly after type I
IFN activation (29). This large potential increased space of cytotoxicity in CD162CD56hi NK cells may be due to their high levels
of miR-378 and miR-30e, which provide large decreasing space
to release many cytolytic effector mRNAs for translation.
It has been reported that mouse NK cells acquiring cytotoxicity
ability after stimulation by activating cytokines requires the
translation of a pre-existing pool of granzyme B and perforin
mRNA (30), indicating that there is a posttranscriptional regulatory mechanism controlling the expression of granzyme B and
perforin expression in mouse NK cells. However, we found that
there is also a posttranscriptional mechanism regulating cytolytic
molecule expression in human NK cells and we demonstrated that
it is through miR-378 and miR-30e action. It is probable that
miRNAs also participate in the posttranscriptional regulation of
cytolytic molecule expression in mouse NK cells, which needs to
be verified in further work. Therefore, possibly in mammals, NK
cells apply this “prearmed” mechanism to ensure rapid responses
to infections or other challenges from both internal and external
sources.
As negative regulators of NK cytotoxicity, miR-378 and miR30e may also participate in the dysfunction of NK cells in NK
cell-associated diseases, such as hepatitis, carcinoma, and various
autoimmunity diseases, which need to be validated in further
researches. Our results (Fig. 4B, 4C) indicated that the posttranscriptional regulation of cytolytic molecules by miR-378 and miR30e is not confined to the process of NK cell activation, but is
universal for activation and function of all cytotoxic cellular
subsets. A study revealed that type I IFN signaling-mediated upregulation of granzyme B expression is required for the enhanced
cytolytic activity of viral-specific CD8+ T cells (31), in which
miR-378 downregulation is probably involved. Malfunction of
these miRNAs may contribute to the irregular expression of cytolytic effectors in CD8+ T lymphocytes during chronic HIV infection (32). Furthermore, the role of these miRNAs in other
cytotoxicity-related clinical disorders is worthy of exploration in
further studies.
In our experiments, we developed a fluorescent labeling tactic
(Cy3-labeled miRNA mimic) to overcome the obstacle of low
transfection efficiency to investigate the functional roles of miRNAs in primary NK cells, which may be of use in other research
for transfection of miRNA or siRNA to primary NK cells. However,
the small amount of successfully transfected cells still limits the
application of this method in functional assays of NK cells.
Based on analysis of miRNome in human NK cells, we found
that type I IFN stimulation enhances cytotoxicity of human NK
cells by downregulating two abundantly expressed miRNAs, miR378 and miR-30e, to enhance cytolytic molecule expression.
During preparation of our manuscript, Fehniger et al. (33) reported
miR-378 AND miR-30e CONTROL NK CELL CYTOTOXICITY
The Journal of Immunology
27.
28.
29.
30.
31.
32.
33.
Malignancies Research Consortium. 2010. Deep sequencing of the small RNA
transcriptome of normal and malignant human B cells identifies hundreds of
novel microRNAs. Blood 116: e118–e127.
Weng, L., X. Wu, H. Gao, B. Mu, X. Li, J. H. Wang, C. Guo, J. M. Jin, Z. Chen,
M. Covarrubias, et al. 2010. MicroRNA profiling of clear cell renal cell carcinoma by whole-genome small RNA deep sequencing of paired frozen and
formalin-fixed, paraffin-embedded tissue specimens. J. Pathol. 222: 41–51.
Kuchenbauer, F., R. D. Morin, B. Argiropoulos, O. I. Petriv, M. Griffith,
M. Heuser, E. Yung, J. Piper, A. Delaney, A. L. Prabhu, et al. 2008. In-depth
characterization of the microRNA transcriptome in a leukemia progression
model. Genome Res. 18: 1787–1797.
Bratke, K., M. Kuepper, B. Bade, J. C. Virchow, Jr., and W. Luttmann. 2005.
Differential expression of human granzymes A, B, and K in natural killer cells
and during CD8+ T cell differentiation in peripheral blood. Eur. J. Immunol. 35:
2608–2616.
Fehniger, T. A., S. F. Cai, X. Cao, A. J. Bredemeyer, R. M. Presti, A. R. French,
and T. J. Ley. 2007. Acquisition of murine NK cell cytotoxicity requires the
translation of a pre-existing pool of granzyme B and perforin mRNAs. Immunity
26: 798–811.
Kohlmeier, J. E., T. Cookenham, A. D. Roberts, S. C. Miller, and
D. L. Woodland. 2010. Type I interferons regulate cytolytic activity of memory
CD8+ T cells in the lung airways during respiratory virus challenge. Immunity
33: 96–105.
Andersson, J., H. Behbahani, J. Lieberman, E. Connick, A. Landay, B. Patterson,
A. Sönnerborg, K. Loré, S. Uccini, and T. E. Fehniger. 1999. Perforin is not coexpressed with granzyme A within cytotoxic granules in CD8 T lymphocytes
present in lymphoid tissue during chronic HIV infection. AIDS 13: 1295–1303.
Fehniger, T. A., T. Wylie, E. Germino, J. W. Leong, V. J. Magrini, S. Koul,
C. R. Keppel, S. E. Schneider, D. C. Koboldt, R. P. Sullivan, et al. 2010. Nextgeneration sequencing identifies the natural killer cell microRNA transcriptome.
Genome Res. 20: 1590–1604.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
17. Kuchen, S., W. Resch, A. Yamane, N. Kuo, Z. Li, T. Chakraborty, L. Wei,
A. Laurence, T. Yasuda, S. Peng, et al. 2010. Regulation of microRNA expression and abundance during lymphopoiesis. Immunity 32: 828–839.
18. Han, Y., M. Zhang, N. Li, T. Chen, Y. Zhang, T. Wan, and X. Cao. 2004. KLRL1,
a novel killer cell lectinlike receptor, inhibits natural killer cell cytotoxicity.
Blood 104: 2858–2866.
19. Griffiths-Jones, S., S. Moxon, M. Marshall, A. Khanna, S. R. Eddy, and
A. Bateman. 2005. Rfam: annotating non-coding RNAs in complete genomes.
Nucleic Acids Res. 33(Database issue): D121–D124.
20. Liu, X., Z. Zhan, D. Li, L. Xu, F. Ma, P. Zhang, H. Yao, and X. Cao. 2011.
Intracellular MHC class II molecules promote TLR-triggered innate immune
responses by maintaining activation of the kinase Btk. Nat. Immunol. 12: 416–
424.
21. Ma, F., S. Xu, X. Liu, Q. Zhang, X. Xu, M. Liu, M. Hua, N. Li, H. Yao, and
X. Cao. 2011. The microRNA miR-29 controls innate and adaptive immune
responses to intracellular bacterial infection by targeting interferon-g. Nat.
Immunol. 12: 861–869.
22. Liu, X., Z. Zhan, L. Xu, F. Ma, D. Li, Z. Guo, N. Li, and X. Cao. 2010.
MicroRNA-148/152 impair innate response and antigen presentation of TLRtriggered dendritic cells by targeting CaMKIIa. J. Immunol. 185: 7244–7251.
23. Li, H., Y. Han, Q. Guo, M. Zhang, and X. Cao. 2009. Cancer-expanded myeloidderived suppressor cells induce anergy of NK cells through membrane-bound
TGF-b1. J. Immunol. 182: 240–249.
24. Ebert, M. S., J. R. Neilson, and P. A. Sharp. 2007. MicroRNA sponges: competitive inhibitors of small RNAs in mammalian cells. Nat. Methods 4: 721–726.
25. Bar, M., S. K. Wyman, B. R. Fritz, J. Qi, K. S. Garg, R. K. Parkin, E. M. Kroh,
A. Bendoraite, P. S. Mitchell, A. M. Nelson, et al. 2008. MicroRNA discovery
and profiling in human embryonic stem cells by deep sequencing of small RNA
libraries. Stem Cells 26: 2496–2505.
26. Jima, D. D., J. Zhang, C. Jacobs, K. L. Richards, C. H. Dunphy, W. W. Choi,
W. Y. Au, G. Srivastava, M. B. Czader, D. A. Rizzieri, et al; Hematologic
221