A 26 parameter, 24 color flow cytometry panel for Human memory NK cell phenotyping

Purpose and appropriate samples

This 26-parameter flow cytometry panel has been developed and optimized to analyze NK cell phenotype, using cryopreserved peripheral blood mononuclear cells (PBMCs) from people living with and without human immunodeficiency virus (PLWH, PWOH). Our panel is designed for the analysis of several parameters of total NK cells and memory NK cell subsets including markers of maturation, activation, and proliferation, as well as activating and inhibitory receptors. Other tissues have not been tested (Table 1).

Background

Natural killer (NK) cells are components of the innate immune system. They are comparable to T cells, in particular CD8 T cells^[1], serving both as cytotoxic effector cells and playing a crucial role in anti-viral and anti-tumor immune responses. However, they differ in recognition, specificity, and memory mechanisms. CD8 T cells use their T cell antigen receptor to recognize the peptide-major histocompatibility complex on the surface of the antigen presenting cells, and subsequently trigger their activation, differentiation, and function. NK cells on the other hand can perform rapid cytolytic and immunomodulatory functions in the absence of prior sensitization. Once activated, NK cells can help clear virus-infected or tumor cells through multiple mechanisms including direct cytotoxicity by releasing perforin, granzyme, immunoregulatory cytokines, and chemokines, or indirectly by influencing adaptive immune responses through their crosstalk with T and dendritic cells^[2–6].

NK cell functional activity is tightly regulated by an array of germline encoded activating and inhibitory receptors. A balance between these receptors determines NK cell activation and responses to alterations due to stress, infections, and cancer. After viral infections, individuals exhibit a reconfiguration of the NK cell receptor repertoire, in particular, an up-regulation of the activating receptor NKG2C and a down-regulation of the inhibitory receptor Siglec-7^[7].

We have designed a 24-color high-dimensional flow cytometry panel (Table 2) that allows us to perform a deep phenotypic analysis of human NK cells. These are defined by the expression of the adhesion molecule CD56 (NCAM-1) and the Fc receptor CD16 (FcγRIIIa), which mediates antibody dependent cellular cytotoxicity (ADCC). Peripheral blood NK cells are mostly mature and cytotoxic and are identified from the lineage-negative cells (CD3⁻ CD19⁻ CD33⁻) as CD56^dim. A smaller proportion of NK cells known as immature or early CD56^bright (Figure 1A) produces more cytokines and chemokines than mature CD56^dim NK cells and are considered their precursors^[8]. A third subset is the dysfunctional CD56^neg NK cell, a cell subtype that lack the expression of CD56 (Figure 1A), have a reduced cytotoxicity, and expand in chronic HIV, Cytomegalovirus (CMV) and Epstein-Barr virus infections^{[9, 10]}.

Table 2.

Reagents used for OMIP

MARKER	FLUOROCHROME	ANTIBODY CLONE	PURPOSE
Live/dead	Aqua		Viability Marker
CD33	BV786	WM53	Non-NK cell lineage exclusion
CD19	BV786	SJ25C1	Non-NK cell lineage exclusion
CD3	BV786	UCHT1	Non-NK cell lineage exclusion
CD16	BUV496	3G8	NK cell subset
CD56	BUV563	B159	NK cell subset
FcεRIγ	FITC	NP_004097	NK Cell subset
CD158a/h/g (KIR2DL1/S1/S3/S5)	PerCP-Cy5.5	HP-MA4	Inhibitory Receptors
CD158e1 (KIR3DL1)	PerCP-Cy5.5	DX9	Inhibitory Receptors
CD328 (Siglec-7)	APC-Vio770	REA214	Inhibitory Receptors
KLRG1	BV421	SA231A2	Inhibitory Receptors
ILT2	BUV615	GHI/75	Inhibitory Receptors
NKG2A (CD159a)	PE-Vio770	REA110	Inhibitory Receptors
CD94	BUV805	HP-3D9	Activating/inhibitory Receptors
NKG2C	PE-Vio615	REA205	Activating Receptors
NKG2D	BB790-P	1D11	Activating Receptors
NKp30	R718	p30–15	Activating Receptors
NKp80	BV650	5D12	Activating Receptors
NKp46	BV711	9E2/NKp46	Activating Receptors
HLA-DR	BV480	G46–6	Activation Marker
CD38	BUV737	HB7	Activation Marker
α4β7	Alexa Fluor 647	Act-1	Homing Marker
Ki67	BV750	B56	Proliferation Marker
CD57	BUV395	NK-1	Maturation Marker
PD-1	BUV661	EH12.1	Exhaustion Marker
EOMES	PE	WD1928	Transcription Factor
T-Bet	PE-Cy5	4B10	Transcription Factor

Figure 1:

Example of staining, gating, and phenotypic analysis of memory NK cells.

A) Identification of NK cells from a representative example of a donor without HIV. First, we verified that there were no irregularities in the acquisition over time (SSC-A vs. Time). From single cells (SSC-H vs. SSC-A) we selected live (Aqua⁻) small lymphocytes (SSC-A vs FSC-A). After exclusion of the lineage-negative cells expressing either CD3/CD19/CD33, total NK cells were identified by the expression of the markers CD56 and CD16. Within the total NK cells, we identified 3 NK cell subsets based on the expression of CD56: early CD56^bright, mature CD56^dim, and the dysfunctional CD56^neg NK cells. We also characterized within the total NK cells the CD57⁺ NKG2C⁺ adaptive memory NK cells and the FcεRIγ⁻ NKG2C⁺ ΔgNK memory cells. B) As the CD57⁺ NKG2C⁺ adaptive memory NK cells and the FcεRIγ⁻ NKG2C⁺ ΔgNK memory cells are overlapping populations, from the total NK cells we define NKG2C+ cells as memory NK cells and subsequently gate the different memory subsets based on the expression of CD57 and FcεRIγ as the following: NKG2C+ CD57+ FcεRIγ-, NKG2C+ CD57+ FcεRIγ+, NKG2C+ CD57- FcεRIγ+, NKG2C+ CD57- FcεRIγ−. C) Representative dot plots of the expression of the parameters included in our panel on total NK cells (i) and NKG2C+ memory NK cells (ii).

Despite being considered innate cells, NK cells can also display features of adaptive immunity. Initial studies reported NKG2C+ NK cells expansion in CMV infection^[11], while later studies identified NKG2C+ memory NK cells with adaptive features in peripheral blood in the context of several other viral infections^[12–15]. These cells were characterized by increased responses to target cells expressing the non-classical human leukocyte antigen E (HLA-E)^{[12, 13]}. Moreover, mounting evidence demonstrates that NK cells have the potential to develop into long-lived and antigen-specific memory cells^[16–19].

Our panel design allows us to analyze some of these memory NK cell subsets. The adaptive memory NK cells are characterized by the acquisition of phenotypic markers CD57 and NKG2C (Figure 1A), and decreased expression of the inhibitory receptor NKG2A^[20]. The FcεRIγ-deficient memory NK cells (ΔgNK) are identifiable by their NKG2C expression, lack of expression of the transmembrane signaling adaptor FcRγ, and downregulation of SYK kinases (Figure 1A). These ΔgNK cells have been shown to exhibit potent antiviral functions against Herpes simplex virus, CMV, HIV, and influenza^[21–23], and exhibit characteristics of long-term persistence and unique epigenetic profiles^{[21, 24]}. They are also less responsive to cytokine stimulation compared to adaptive CD57+NKG2C+ memory NK cells^[24]. On the other hand, they are associated with enhanced ability to mediate ADCC^[21–23], a feature that could be highly beneficial in the context of vaccination. More recently, the transmembrane signaling adaptor FcRγ-deficient memory NK cells were associated with lower parasitemia and resistance to malaria^[15].

Our panel design included several inhibitory and activating receptors relevant for NK cell functionality and differentiation. We included markers to analyze the expression of inhibitory receptors such as the heterodimer CD94/NKG2A and CD158 killer immunoglobulin-like receptors (KIR). These are involved in NK cell inhibition and maintenance of self-tolerance through interaction with the non-classical human leukocyte antigen E (HLA-E)^[25] and the classical major histocompatibility complex molecules, respectively^{[3, 26, 27]}. We have included the CD158e1 (KIR3DL1) and the CD158a/h/g-specific clone HP-MA4 which recognizes several CD158 proteins knowns as KIR2D, specifically KIR2DL1 (CD158a), KIR2DS1 (CD158h), KIR2DS3, and KIR2DS5 (CD158g). However, it has been shown that a minority of individuals could be negative for these CD158 proteins^{[28, 29]}.

We have included the adhesion inhibitory receptor sialic acid-binding immunoglobulin-like lectin 7 (Siglec-7), which is mostly expressed by highly functional mature NK cells^[30]. The decreased expression of Siglec-7 has been described as an early marker of dysfunctional NK cells in HIV infection, preceding the down-modulation of CD56^[31]. Another inhibitory receptor we added to our panel is ILT-2, also known as LILRB1 and CD85j. ILT-2, binds to MHC-I as well as non-classical MHC-I molecules, such as HLA-F and HLA-G. Its ligation inhibits NK cell activation and proliferation^[32–34], and it has been shown that blocking ILT-2 restores NK cells function in chronic lymphocytic leukemia.

Our panel allows the analysis of Natural cytotoxicity receptors (NCRs) such as NKp30, NKp46, and the activating co-receptor NKp80. These receptors play a central role in triggering NK cell activation and are expressed in both resting and activated cells. An increased expression of these NCRs has been associated with a reduced HIV-1 reservoir size^[35], while a decreased expression leads to impaired NK cell cytotoxicity^[36]. The activating co-receptor NKp80 mediates the activation of NK cells and its expression has been shown to be reduced in people living with HIV (PLWH)^[37]. We also included receptors from the NKG2 family, which are predominantly expressed on NK cells and a subset of CD8 T cells^[38]. NKG2A and NKG2C form a heterodimer with CD94^[38], and NKG2D rather associates with itself and forms a homodimer ^[39]. CD94/NKG2A is highly expressed by early NK cells, where engagement to its non-classical HLA-E ligand leads to NK cell inhibition. In contrast, the activating receptor NKG2C is expressed on mature CD56^dim NK cells and enables the identification of two distinct but overlapping memory subsets. The adaptive CMV-memory NK cells are characterized by co-expression of NKG2C and CD57, while ΔgNK memory cells are also NKG2C+ but lack expression of the transmembrane signaling adaptor FcRγ (Figure 1A). Both of these memory subsets exhibit adaptive immune features such as clonal-like expansion and long-term persistence^[40]. As these two memory subsets are overlapping populations (see online Fig 5), we analyzed memory NK cells as a combination of ΔgNK and adaptive memory NK cells phenotypes. We then identify NKG2C+ cells as memory NK cells and subsequently based on the expression of CD57 and FcεRIγ, we describe 4 memory subsets: NKG2C+ CD57+ FcεRIγ-, NKG2C+ CD57+ FcεRIγ+, NKG2C+ CD57- FcεRIγ+, and NKG2C+ CD57- FcεRIγ- (Figure 1B).

We were interested in analyzing the activating receptor NKG2D, having been shown that its ligand engagement co-stimulates CD16 signaling and enhances ADCC mediated killing of HIV-infected cells, but its expression is reduced in chronic HIV infection^[41]. Due to our interest in the role of memory NK cells in HIV infection, the inhibitory receptor KLRG1 was also included after being recently shown that its expression is increased on antigen-specific memory NK cells^[18].

Viral infections such as HIV can result in chronic immune activation, leading to a persistent increased activation of NK cells despite viral suppression post-antiretroviral treatment ^{[42, 43]}. To analyze NK cell activation, markers such as HLA-DR and CD38 were added to our panel. We have also included the checkpoint inhibitory receptor PD-1, as its expression has been reported on NK cells in HIV and CMV infections as well as in several tumor models. Checkpoint receptor blockade of PD-1 has also been shown to reverse NK cell impairment^{[44, 45]}.

We also included in our panel the transcription factors T-box expressed in T cells (T-bet) and eomesodermin (Eomes) which play a crucial role in NK cell development and effector potential. Both are modulated during maturation of NK cells, with progressive T-bet upregulation and Eomes downregulation toward terminal differentiation^[46].

We added the gut-homing marker α4β7 to our analysis, as NK cells undergo large shifts in trafficking to lymph nodes and gut mucosa in response to viral infections. An expansion of CD56+ α4β7+ cytotoxic NK cells was observed in response to SIV infection ^{[47, 48]}. Lastly, we have also included the proliferation marker Ki67, increased expression of which has been observed in the dysfunctional CD56^neg NK cells leading to their accumulation in PLWH.

Similarity to published OMIPs

Several previously published OMIPs such as OMIP-007, 027, 029, 039, 058, 064, 070, and 080 include NK cell markers, enabling the characterization of NK cells. However, our panel allows us to perform a thorough and deep analysis of NK cells, including of all 3 subsets based on CD56 expression, and the memory subsets which are expanded in chronic infections such HIV.

Our panel covers over 16 immune parameters that are altered in NK cells during viral infections, and would be particularly useful for analysis of rare and small aliquot specimens from unique human cohorts.

Table 1.

Summary table for OMIP

Purpose	Phenotypic profile analysis of total NK cells and memory subsets
Species	Human
Cell types	PBMCs
Cross-References	OMIP-007, 027, 029, 039, 058, 064, 070 & 080