Avaliação de aspectos inatos e adaptativos do sistema imune na psoríase: análise fenotípica e... por Mariana Dias Batista - Versão HTML

ATENÇÃO: Esta é apenas uma visualização em HTML e alguns elementos como links e números de página podem estar incorretos.
Faça o download do livro em PDF, ePub, Kindle para obter uma versão completa.

Ag a cél. T

Fumo

Células

dendríticas

naive

plasmocitóides

Migração de

liberam IFN-α

células

IL-12

dendríticas

CXCR3

Migração de

plasmocitóides

Diferenciação

outras células T

em resposta a

Th-1/Tc-1

IFN-γ

Processo

quemerina

Células

inflamatório

CLA

dendríticas

Células NK e

mielóides

NKT liberam

ativadas

IFN-γ

Imunidade Inata

Imunidade Adaptativa

Figura 1.1 Mecanismos patogenéticos na psoríase. Indivíduos geneticamente

predispostos, quando expostos a fatores desencadeantes, iniciam processo de ativação

dos queratinócitos, que culmina com a liberação de IFN-" por células dendríticas

plasmocitóides, que por sua vez estimulam células dendríticas mielóides a apresentar

antígeno ainda não determinado para células T naïve. Estas por sua vez, sob estímulo

de IL-12, se diferenciam para a via Th-1, com expressão de CXCR3 e CLA, e migram

para a pele. Por outro lado, células T naïve sob o estímulo de IL-23 se diferenciam em

células Th-17, que expressam CLA, CCR6 e CCR4 e também contribuem para a

patogênese da doença. Adaptado de: Nestle, 2009 e Nograles, 2008.

16

2 OBJETIVOS

2.1 Objetivos gerais

2.1.1 Investigar se o fenótipo de células NK e células T na psoríase é compatível com

imuno-senescência e maior capacidade efetora

2.1.2 Avaliar características funcionais de células T na psoríase

2.2 Objetivos específicos

2.2.1 Avaliar a expressão do marcador de imunosenescência CD57, do marcador

inibitório de células NK NKG2A e do marcador ativador de células NK NKG2C em

células T e células NK isoladas de pele lesional, não afetada e células mononucleares

de sangue periférico (CMSP) de pacientes com psoríase, comparadas àquelas isoladas

de pele e CMSP de controles sadios

2.2.2 Avaliar a produção das citocinas IL-17A, IL-22, IL-2, TNF-", IFN-#, IL-21, IL-

27 e IL-33 por células T CD4+ e CD8+ isoladas da pele lesional e não afetada de

pacientes com psoríase, após expansão em cultura e estímulo com mitógenos.

17

3 MÉTODOS

3.1 Casuística

Os pacientes foram recrutados no Serviço de Dermatologia do San Francisco General

Hospital, na Universidade da California, São Francisco, após aprovação pelo Comitê de

Ética em Pesquisa daquela instituição e da Universidade de São Paulo (anexos A e B).

Todos os pacientes participantes do estudo preencheram o Termo de Consentimento

Livre e Esclarecido (anexo C). Na primeira fase de recrutamento, foram incluídos 12

pacientes (P1 a P12), que participaram da porção fenotípica do estudo. Em uma fase

subsequente, foram incluídos outros 13 pacientes (F1 a F13), que participaram da

porção de ensaios funcionais de células T, conforme demonstra a tabela 3.1.

Tabela 3.1- Características amostrais e demográficas dos pacientes

Tempo desde o

Subtipo de

diagnóstico

Paciente

Sexo

Idade

psoríase

Gravidade

Amostras Coletadas

Fatores Predisponentes

(anos)

P1

M

22

vulgar

moderada

pele lesional/ pele não afetada/ sangue

estresse/ trauma cutâneo

10

P2

F

35

gutata

moderada

pele lesional/ pele não afetada/ sangue

nenhum

27

P3

M

61

vulgar

moderada

sangue

trauma cutâneo/ inverno

3

P4

F

84

gutata

grave

sangue

estresse/ inverno

20

P5

M

67

gutata

leve

pele lesional/ pele não afetada/ sangue

nenhum

7

P6

M

44

gutata

leve

pele lesional/ pele não afetada

estresse

24

P7

M

21

gutata

leve

pele lesional/ pele não afetada/ sangue

inverno

1

P8

F

41

vulgar

leve

pele lesional/ pele não afetada/ sangue

estresse/ consumo de álcool/ inverno

11

P9

M

37

gutata

leve

pele lesional/ pele não afetada/ sangue

inverno

1

P10

M

67

vulgar

leve

pele lesional/ pele não afetada/ sangue

nenhum

9

P11

M

52

gutata

leve

pele lesional/ pele não afetada/ sangue

nenhum

40

P12

M

61

vulgar

leve

pele lesional/ pele não afetada/ sangue

estresse

10

F1

M

50

vulgar

grave

pele lesional/ pele não afetada

estresse

5

F2

F

53

vulgar

leve

pele lesional/ pele não afetada

estresse/ inverno

38

F3

M

68

vulgar

leve

pele lesional/ pele não afetada

estresse

3

F4

F

53

vulgar

leve

pele lesional/ pele não afetada

trauma cutâneo/ inverno

50

F5

M

40

vulgar

moderada

pele lesional/ pele não afetada

estresse/ inverno/ trauma cutâneo

12

F6

M

46

vulgar

moderada

pele lesional/ pele não afetada

nenhum

14

F7*

F

54

vulgar

moderada

pele lesional/ pele não afetada

estresse/ inverno

12

F8

F

27

vulgar

grave

pele lesional/ pele não afetada/ sangue

estresse/ inverno/ gravidez

8

F9

M

60

vulgar

moderada

pele lesional/ pele não afetada/ sangue

nenhum

42

F10

F

58

vulgar

moderada

pele lesional/ pele não afetada/ sangue

inverno

37

F11

F

48

vulgar

leve

pele lesional/ pele não afetada

estresse/ inverno/ período pré-menstrual

32

F12

F

55

vulgar

grave

pele lesional/ pele não afetada

estresse/ trauma cutâneo

50

F13

F

32

vulgar

grave

pele lesional/ pele não afetada

estresse

19

Os pacientes apresentavam diagnóstico de psoríase em placas ou gutata, com

gravidade classificada com base na área de superfície corporal afetada ( body surface

affected area- BSA), da seguinte maneira: leve- menos de 5%, moderada- 5 a 30%,

grave- mais de 30%. Os pacientes tinham o diagnóstico de psoríase há pelo menos 1

ano, estavam sem tratamento tópico ou sistêmico há pelo menos 6 meses, e foram

recrutados durante sua primeira consulta no Serviço de Dermatologia. Foram

excluídos pacientes que apresentassem outras doenças auto-imunes e infecção pelo

index-29_1.png

18

HIV, ou que estivessem em qualquer tipo de tratamento tópico ou sistêmico para a

psoríase. Foram coletadas amostras de sangue, pele lesional e pele não afetada dos

pacientes. Para a porção fenotípica do estudo, foi possível coletar amostras pareadas

de pele lesional, não afetada e sangue de 9 dos 12 pacientes. Para os 3 outros

pacientes, coletou-se apenas sangue ou pele, conforme indica a tabela 1. Quando na

citometria os números de células identificados após o gating eram muito pequenos

(abaixo de 30 eventos) os pacientes foram excluídos das análises, conforme

demonstram as figuras nos resultados. Na porção funcional do estudo, foram

coletadas amostras de pele lesional e não afetada de 13 pacientes, e amostras pareadas

de sangue de 3 (tabela 3.1). Foi excluído um paciente da porção funcional do estudo

(F7) devido a baixa pureza após o sorting celular.

As amostras de sangue de 10 controles sadios foram obtidas após consentimento no

Banco de Sangue de Stanford. A mediana de idade dos controles foi de 45 anos,

conforme demonstra a tabela 3.2. As amostras de pele normal de 3 controles sadios

foram obtidas de tecido descartado após abdominoplastia, em pacientes sem histórico

de psoríase ou outras afecções inflamatórias.

3.2 Processamento de amostras de pele

As amostras de pele foram coletadas através de biópsia com punch de 4mm, uma

realizada na borda de lesão ativa de psoríase, e outra em pele não afetada localizada a

pelo menos 5 cm de distância das lesões ativas. As amostras foram incubadas em

RPMI-1640 com colagenase/dispase na concentração de 1 mg/ml (Roche Diagnostics,

Indianapolis, IN, USA), associadas a 10 unidades/ml de DNAse I recombinante

index-30_1.png

index-30_2.png

index-30_3.png

index-30_4.png

index-30_5.png

index-30_6.png

19

(Roche Diagnostics, Indianapolis, IN, USA), a 37°C por 15 minutos, e posteriormente

a 4°C durante 8 horas. Após este período, foi realizada a separação mecânica entre a

epiderme e a derme, e a epiderme foi tratada com tripsina e EDTA (GIBCO

Invitrogen, Carlsbad, CA, USA) em tampão fosfato (PBS), a 37°C por 15 minutos.

Para determinar se havia necessidade de cultura para isolar as células linfo-

monocitárias da pele, após a separação entre epiderme e derme, foi realizado

experimento com amostras de pele normal proveniente de abdominoplastia, no qual se

comparou células provenientes da etapa de digestão com colagenase sem tempo

nenhum de cultura com outros dois grupos: células cultivadas por 48 horas e por 7

dias. Conforme demonstra a figura 3.1, determinou-se que a melhor viabilidade e

separação de células linfo-monocitárias se deu no período de 48 horas de cultura.

250K

250K

250K

200K

200K

200K

150K 83.8

150K

150K

2.81

43.2

SSC-A

SSC-A

SSC-A

100K

100K

100K

50K

50K

50K

0

0

0

0 102

103

104

105

0

103

104

105

0 102

103

104

105

<AARD-A>: LIVE/DEAD

<AARD-A>

<AARD-A>: live/dead

Figura 3.1 – Determinação do tempo de cultura das células isoladas da pele. A-

Citometria realizada a fresco sem cultura. B- 48 horas de cultura. C- 7 dias de cultura

Assim, nas amostras dos pacientes, após a digestão enzimática com colagenase, a

epiderme e a derme foram subsequentemente cultivadas separadamente a 37°C por 48

horas, em RPMI-1640 suplementado com soro humano a 10% (Gemini Bio Products,

Sacramento, CA, USA), penicilina e estreptomicina a 1%, e tampão HEPES a 1 mol/L

(Invitrogen, Carlsbad, CA, USA), em placas de cultura de 6 poços. Após esse período,

as suspensões celulares foram obtidas filtrando o meio de cultura em filtro de 70 µm

(BD, San Jose, CA, USA). Devido aos pequenos números de células obtidos em

contagem no hemocitômetro, as células isoladas da epiderme e da derme foram

combinadas para produzirem números suficientes para a realização de citometria de

20

fluxo na porção fenotípica do estudo, e para sorting de células T CD4+ e CD8+ na

porção funcional do estudo.

Com o objetivo de determinar se o tratamento com colagenase/dispase poderia alterar

o processo de coloração com os fluorocromos para citometria de fluxo, foi realizado

experimento no qual se comparou CMSP tratadas com colagenase/dispase com outras

que não receberam este tratamento, e após citometria de fluxo não houve diferença

entre os grupos.

Foi também realizado experimento piloto com amostra proveniente de

abdominoplastia para determinar se a expressão de receptores observada em células

isoladas a partir de punch de 4mm era representativa da expressão observada após

digestão enzimática de amostra maior, de 3x3 cm de pele. Não foi observada

diferença entre as amostras, sugerindo que o punch de 4mm foi representativo do

ambiente inflamatório local.

3.3 Processamento de amostras de sangue

As amostras de sangue foram coletadas em tubos vacutainer com EDTA, e as CMSP

foram isoladas através do gradiente de Ficoll-Hypaque, no período de até 6 horas após

a coleta. Após a separação, as CMSP foram lavadas 2 vezes com PBS, e re-suspensas

em tampão FACS para citometria de fluxo (PBS com albumina bovina 0.5% e 2mM

EDTA).

3.4 Citometria de fluxo

Para a porção fenotípica do estudo, após o isolamento das suspensões celulares, foi

realizada fenotipagem para citometria de fluxo. Primeiramente, os receptores Fc nas

células foram bloqueados com IgG humana 10 µg/ml (Sigma Aldrich, S. Louis, MI,

USA), durante 20 minutos no gelo. Posteriormente, as células foram marcadas com

anticorpos monoclonais, descritos na tabela 3.3, por 30 minutos no gelo, depois

lavadas duas vezes com tampão FACS, e fixadas com paraformaldeído a 2%

(Touisimis, Rockville, MD) em PBS. As amostras foram adquiridas no citômetro de

fluxo LSR II (BD Biosciences). Por fim, os dados gerados foram analisados através

do software FlowJo versão 8.8.6 (Tree Star, San Carlos, CA, USA).

index-32_1.png

21

As janelas de aquisição ( gating) utilizadas estão representadas na figura 3.2.

Primeiramente, foram selecionadas as células únicas (área e altura do tamanho

celular), em seguida foi utilizada a janela de aquisição baseada na granularidade e

tamanho celular. Posteriormente, foram excluídas células mortas através de seleção de

células Amine-Acqua negativas, e novamente por tamanho e granularidade

identificou-se a população de linfócitos. Para identificação mais fidedigna dos

linfócitos, foram selecionadas as células CD45+. Posteriormente, em gráfico de CD3

X CD14/CD19, foram selecionadas 2 subpopulações: CD3+CD14-CD19-, para

células T, que foram então subdivididas em CD4+ e CD8+, e CD3-CD14-CD19-, para

células NK, que foram então subdivididas de acordo com a expressão de CD56 e

CD16. Os marcadores CD14 e CD19 foram utilizados para excluir monócitos e

linfócitos B, respectivamente.

index-33_1.png

22

Figura 3.2- Estratégia de gating para citometria de fluxo

3.5 Sorting de células T CD4+ e CD8+

Após o isolamento a fresco de CMSP e de células provenientes das amostras de pele,

foi realizada marcação com anticorpos monoclonais, conforme indica a tabela 3.4, por

30 minutos no gelo. As células foram lavadas duas vezes com tampão FACS, e foi

realizado sorting celular no citômetro FACS Aria (BD Biosciences, San Jose, CA,

USA). A estratégia de aquisição realizada envolveu seleção de células únicas, células

CD45+, CD3+CD4+ ou CD3+CD8+. No primeiro experimento (n=6), não houve

viabilidade das células T CD3+CD4+ após o sorting, e as etapas subsequentes foram

realizadas apenas com as células T CD3+CD8+. No segundo experimento (n=7),

células T CD3+CD4+ e CD3+CD8+ foram isoladas e utilizadas nas etapas

subsequentes. A pureza de todos os procedimentos de sorting foi superior a 95%,

index-34_1.png

23

exceto para o paciente F7, cujas amostras obtiveram pureza de 75% e foram excluídas

das etapas subsequentes.

3.6 Cultura de células T

Células T CD4+ e CD8+ isoladas de pele lesional, não afetada e sangue de pacientes

com psoríase foram cultivadas por 8 horas a 37°C e 5% CO2, numa placa de cultura

de 96 poços, fundo em U. Cada amostra foi dividida em dois, e a metade das células

foi estimulada com acetato de forbol miristato (PMA) (50ng/ml) e ionomicina

(500ng/ml) (Sigma-Aldrich, St. Louis, MO, USA), e a outra metade das células não

recebeu a estimulação. Após o período de 8 horas, a placa foi centrifugada e os

sobrenadantes foram retirados. As células foram lavadas com PBS, e meio de cultura

contendo IL-2 400 UI/ml foi adicionado. As células foram então para cultura por 7

dias a 37°C e 5% CO2. No sétimo dia as placas foram centrifugadas e os

sobrenadantes foram congelados a -20°C para uso subsequente nos ensaios de

multiplex.

3.7 Ensaios de Multiplex

A produção de IFN%#, TNF%", IL-2, IL-17A, IL-22, IL-21, IL-27 e IL-33 foi avaliada

através do ensaio de multiplex para citocinas denominado Legendplex (Biolegend,

San Diego, CA, USA). Foi montada placa sob encomenda para o ensaio, que continha

todas as citocinas simultaneamente. A presença das citocinas foi avaliada nos

sobrenadantes coletados após a cultura de células T CD4+ e CD8+ previamente

isoladas por sorting. As amostras foram adquiridas no analizador Labscan 200

(Luminex, Austin, TX, USA), utilizando o software Bio-Plex manager versão 6.0

(Bio-Rad, Hercules, CA, USA).

24

3.8 Análise Estatística

As análises estatísticas foram realizadas no software Prism versão 4.0a (GraphPad, La

Jolla, CA, USA). As análises foram realizadas através do teste não-paramétrico de

Mann-Whitney quando se comparavam 2 grupos, ou do teste não paramétrico de

Kruskal-Wallis quando se comparavam 3 grupos. O teste não-paramétrico de

Wilcoxon foi utilizado nas análises de amostras pareadas. Foi utilizado valor mínimo

de significância estatística de p=0.05. O software SPICE versão 5.2 (Exon NIAID,

Bethesda, MD, USA) foi utilizado para as análises de co-expressão realizadas na

porção fenotípica do estudo.

25

4 RESULTADOS

4.1 Manuscrito de artigo científico sobre células NK

4.1.1 Folha de Rosto

Title: Skewed Distribution of Natural Killer Cells in Psoriatic Skin Lesions

Running Head: NK Cells in Psoriatic Skin Lesions

Authors: M. D. Batista, MD1,2, E. L. Ho, MD, PhD1#, P. J. Kuebler, PharmD1, J. M.

Milush, PhD1, L. L. Lanier, PhD3, E. G. Kallas, MD, PhD2,4, V. A. York, BS1, D.

Chang, MD5, W. Liao, MD6, P. Unemori, MD6, K. S. Leslie, M.B.,B.S., FRCP6, T.

Maurer, MD6, D. F. Nixon, MD, PhD1*

Affiliations: 1Division of Experimental Medicine, Department of Medicine,

University of California San Francisco, San Francisco, California, USA, 2Division of

Clinical Immunology and Allergy, School of Medicine, University of São Paulo, São

Paulo, Brazil, 3Department of Microbiology and Immunology and the Cancer

Research Institute, University of California, San Francisco, San Francisco, USA,

4Instituto de Investigação em Imunologia, University of São Paulo, São Paulo, Brazil,

5California Pacific Medical Center - Davies Campus, San Francisco, CA, USA,

6Department of Dermatology, University of California, San Francisco, USA. #

Current Address: Department of Neurology, University of Washington, Seattle, WA.

O presente manuscrito foi submetido e encontra-se em revisão no periódico

Experimental Dermatology (Anexo D).

26

4.1.2 Abstract

Background: Psoriasis is a hyper-proliferative inflammatory disease of the skin in

which immunological mechanisms play a direct role in disease pathogenesis. There

have been limited studies of natural killer (NK) cells in psoriasis.

Objectives: To examine the phenotype and functional role of NK cells in skin biopsies

and peripheral blood mononuclear cells from patients with psoriasis and healthy

controls.

Methods: CD56+CD16- and CD56+CD16+ NK cells were isolated from lesional skin,

unaffected skin and PBMC of psoriasis patients, and normal skin and PBMC from

healthy controls. The expression of CD57, NKG2A, and NKG2C was assessed by

flow cytometry.

Results: NK cells in psoriatic skin lesions were skewed in their expression of CD57, a

marker of NK cell maturity, with CD57 expression significantly reduced and NKG2A

expression increased on NK cells in lesional and unaffected skin compared to

controls.

Conclusions: These data suggest that NK cells in psoriasis lesions exhibit an

immature phenotype, but further studies are needed to determine the functional role of

these cells in psoriasis.

Key Words: psoriasis, natural killer cells, CD57, immunosenescence

27

4.1.3 Introduction

Psoriasis is a chronic disease of the skin with significant morbidity, where clinical

improvements can be observed with the use of different therapeutic modalities

including topical and systemic immunosuppressive agents, anti-cytokine biological

therapies, and ultraviolet light therapies. Histologically, psoriasis is characterized by

psoriasiform epidermal hyperplasia, parakeratosis, loss of the granular cell layer, and

the formation of spongiform pustules (1).

The etiologic cause of psoriasis is unknown, although certain environmental triggers

(e.g. infection, skin injury, stress, weather, and medications) are thought to contribute

to its onset. The development of psoriatic lesions is thought to be due to the

interaction of these triggers in persons with genetic predisposition (1). The major

genetic determinant of psoriasis is located within the MHC region on chromosome 6,

in a locus known as PSORS1 (2). From this region, HLA-C*06:02 allele is most

strongly associated with psoriasis (3). Considerable genetic variability can be

observed among different subtypes of psoriasis, with guttate psoriasis being most

associated with PSORS1 (4). However, the underlying cellular and molecular

mechanisms leading to the hyper-proliferative inflammatory nature of psoriasis

remain incompletely understood.

Several studies have characterized the cellular and cytokine immune composition of

healthy and psoriatic skin. These studies suggest contributions from both the innate

and adaptive immune systems and their interaction with keratinocytes and other skin

cells in the pathogenesis of psoriasis (5-7). Increasingly, evidence points to a pro-

inflammatory environment with local expansion of activated lymphocytes within

psoriatic lesions, elevated IFN-#, TNF-", IL-17, IL-22, and IL-23, as well as secretion

of antimicrobial peptides such as cathelicidin from keratinocytes (8-10). Although IL-

23 may contribute to disease, a role for IL-12 in psoriasis is also described (11). IL-17

and IL-22 co-producing T cells are also considered important contributors to the pro-

inflammatory milieu associated with psoriasis pathology (8, 12, 13).

28

Natural Killer (NK) cells are lymphocytes at the interface of the innate and adaptive

immune systems. They migrate into sites of infected and damaged tissue in response

to elevated levels of pro-inflammatory chemokines and can further produce

significant amounts of pro-inflammatory cytokines, such as IFN-#. While innate

immune mechanisms may also play a role in the immunopathogenesis of psoriasis, the

role of NK cells in psoriasis is less studied (14-16). The interaction between HLA-C

and Killer cell Immunoglobulin-like Receptors (KIRs) on NK cells has been

investigated in psoriasis, and KIR2DS1 is associated with psoriasis vulgaris and

psoriatic arthritis (17-19). Furthermore, greater expression of HLA-G, a known

inhibitor of NK cell cytolysis, is observed in the skin of psoriasis patients compared to

normal skin (20, 21).

NK cells express a heterogeneous repertoire of germline-encoded receptors that serve

to educate them through an appropriate balance of activating and inhibitory signals as

they mature (22). These receptors also regulate NK cell effector function (23).

Myeloid dendritic cells can influence NK cell function through cytokine secretion

(24), particularly IL-12, which is a potent inducer of IFN-# production by NK cells.

We investigated the expression of the inhibitory receptor NKG2A and the activating

receptor NKG2C, which have a role in immunosurveillance by binding to HLA-E

(25), on NK cells in the skin and blood of patients with psoriasis and healthy controls.

Although both NKG2A and NKG2C bind HLA-E, NKG2C binds with a lower

affinity than NKG2A (25, 26). However, as NK cells mature, they tend to express

more NKG2C and lose NKG2A expression (27).

Although NK cell function (cytotoxicity and cytokine secretion) is related to their

stage of development, both immature CD56brightCD16neg and mature

CD56dimCD16+ NK cell subpopulations secrete IFN-#, which is critical in the

pathogenesis of psoriasis (28). Indeed, one study demonstrated how specific target

cell ligands dictate the qualitative and temporal aspects of NK cell cytokine responses

resulting in graded responses, depending on the multiplicity of activating receptors

engaged. That study suggested the CD56dimCD16+ NK cell subpopulation may be

the most important source of cytokines upon recognition of aberrant target cells (29).

Expression of CD57 identifies terminally differentiated T cells (30). Recently, we,

29

and others, have shown that CD57 is also expressed on highly mature cells within the

CD56dimCD16+ NK cell compartment, became NKG2Chi, and finally acquired

CD57 (31, 32). This suggests that CD57 might provide a marker of "memory" NK

cells (27). These CD57+ NK cells have a poor replication capacity in vitro, but a high

capacity to produce IFN-# following stimulation through the activating receptor CD16

(31, 32). More importantly, CD57+CD56dimCD16+ NK cells are less responsive to

stimulation

by

IL-12

and

IL-18

compared

to

their

less

mature

CD57negCD56dimCD16+ counterparts (32). These data suggest that infiltration by

less mature CD57negCD56dimCD16+ NK cells may fuel the pro-inflammatory

environment as they are more capable of responding to activating cytokines.

In view of our finding that CD57 expression can identify NK cell subsets capable of

producing high amounts of IFN-# in response to pro-inflammatory cytokines (32), and

our observation that a significant amount of the total IFN-# produced in response to

antigenic stimulation in a polyclonal PBMC population can be attributed to NK cells

rather than T cells (33), we examined the phenotype of NK cells in psoriasis by

investigating the distribution of CD57+ NK cells in lesional and non-lesional skin, as

well as peripheral blood of patients with psoriasis compared to controls.

30

4.1.4 Methods

Study design and skin samples

Peripheral blood mononuclear cells (PBMCs) and skin biopsies were collected from

patients with psoriasis (lesional and non-lesional skin) (n=12) and PBMC alone were

collected from healthy controls (n=10). Normal skin from healthy donors was

obtained from discarded tissue following plastic surgery procedures (n=3). All

patients gave informed consent and the study was approved by the UCSF Institutional

Review Board. Patients with mild to severe psoriasis of both guttate or plaque subtype

were included. Severity was assessed based on affected body surface area (BSA), as

follows: mild - less than 5% BSA, moderate 5 - 30% BSA, severe - more than 30%

BSA. Additionally, patients had a history of psoriasis for at least 1 year and were

treatment-free in the last 6 months. Patients undergoing any topical or systemic

therapy were excluded.

Skin sample preparation

Tissue samples from psoriasis patients were collected as a 4-mm punch biopsy. One

punch biopsy was collected from an active psoriasis lesion (close to the lesion

margin) and another from non-lesional skin, at least 5 cm away from an active lesion.

Tissue samples were incubated overnight at 4°C in 1 mg/ml collagenase and dispase

(Roche Diagnostics, Indianapolis, IN, USA) and 10 units/ml recombinant DNAse I

(Roche Diagnostics, Indianapolis, IN, USA) in RPMI-1640. The epidermis was

subsequently treated with 0.25% trypsin and EDTA (GIBCO Invitrogen, Carlsbad,

CA, USA) in PBS at 37°C for 15 minutes. Epidermis and dermis were both cultured

separately for 48 hours at 37°C in RPMI-1640 supplemented with 10% pooled human

serum (Gemini Bio Products, Sacramento, CA, USA), 1% penicillin and

streptomycin, and 1 mol/L HEPES buffer (Invitrogen, Carlsbad, CA, USA) in 6-well

culture plates. Single cell suspensions were obtained after rinsing through a 70 µm

cell strainer (BD, San Jose, CA, USA), and cells isolated from epidermis and dermis

were subsequently combined to produce sufficient cell numbers for flow cytometry.

We also treated PBMC with collagenase and dispase to determine whether NK cell

staining within PBMC was perturbed by this treatment. No differences were observed

(data not shown).

31

Flow cytometry

Twelve-parameter flow cytometry was performed using a LSR II flow cytometer (BD

Biosciences). Fc receptors on cells were first blocked using 10 µg/ml human IgG

(Sigma Aldrich, S. Louis, MI, USA) for 20 minutes on ice. Cells were then stained for

30 minutes on ice with fluorophore-labeled antibodies, washed with FACS buffer

containing PBS, 0.5% bovine serum albumin, and 2mm EDTA, and fixed with 2%

paraformaldehyde in PBS. The data files were analyzed using FlowJo Software

version 8.8.6 (Tree Star, San Carlos, CA, USA).

Statistical Analyses

The statistical analyses were performed using Prism Software version 4.0a

(GraphPad, La Jolla, CA, USA). Groups were compared using the Mann-Whitney

non-parametric test with a minimum significance value of p = 0.05. SPICE software

version 5.2 was used for the analyses presented in Figure 2.

32

4.1.5 Results

Patient Demographics

Twelve subjects with psoriasis were studied, including 9 males and 3 females. The

median age was 48 years (range 21 to 84). With respect to psoriasis subtype, 5

subjects had plaque psoriasis and 7 subjects had guttate psoriasis. The majority of

subjects had mild disease (n = 8), 3 subjects had moderate disease, and 1 subject had

severe disease. Matched skin and PBMC samples were available from 9 out of 12

psoriasis subjects. Healthy control PBMC samples were available for 10 individuals,

with a median age of 45 years. In patients for whom serum was available, CMV

serology was positive. This positivity is representative of the general demographic of

the local population.

Distribution and phenotype of NK cells in lesional and non-lesional skin and

peripheral blood of psoriasis patients compared to healthy controls

We examined the distribution and phenotype of NK cells in skin and peripheral blood

of psoriasis patients and healthy controls. Lymphocytes were identified by forward

and side scatter followed by staining with CD45. Within the CD45+ lymphocyte

subset, both lesional and non-lesional skin of psoriatic patients had similar

frequencies of CD56+CD16+ NK cells; however, in the blood there was a trend toward

an increased frequency of CD56+CD16+ NK cells in patients compared to healthy

controls (Fig. 1a). To assess the maturity of NK cells in the skin and peripheral blood,

we measured CD57 expression on CD56+CD16+ NK cells. The frequency of CD57+

NK cells in lesional skin (less than 10%) was significantly lower than in non-lesional

skin (~ 40%) (Fig. 1b). Interestingly, the non-lesional skin in psoriasis subjects had

significantly fewer CD57+ NK cells than in skin of healthy controls. This difference

was specific to the skin, as no significant difference in CD57+CD56+CD16+ NK cells

was observed in the blood of patients compared to healthy controls (Fig. 1b).

An assessment of inhibitory, NKG2A, and activating, NKG2C, receptor expression

on skin and peripheral blood CD56+CD16+ NK cells revealed a significant expansion

of NKG2A+ NK cells in lesional and non-lesional skin of patients compared to

healthy controls (Fig. 1c). However, the frequency of NKG2A+ NK cells in the

peripheral blood was similar between psoriasis patients and healthy controls. In

contrast, the proportion of NKG2C+ NK cells was similar in both lesional and non-

lesional skin to healthy controls. In the blood, psoriasis patients had a significantly

33

greater frequency of NKG2C+ NK cells compared to healthy controls of similar

median age (Fig. 1d).

When we examined co-expression of CD57, NKG2A, and NKG2C on skin NK cells

we observed a marked difference between normal skin from healthy donors and

lesional and non-lesional skin from psoriasis patients (Fig. 2). The majority of NK

cells from the skin of psoriasis patients only expressed NKG2A, particularly in

lesional skin. In contrast, the majority of NK cells in the skin of healthy donors

contained CD57+ NK cells that were NKG2A- and NKG2C-. These data suggest NK

cells in the skin of psoriasis patients, particularly in the lesion itself, are less mature

NK cells. When we compared plaque and guttate psoriasis for all the measured

parameters, no difference was found (Supplemental Fig. 1).

index-45_1.png

34

Figure 1: (a) Frequency of CD56+CD16+ NK cells in lesional skin (n=7) and

unaffected skin (n=8) of psoriasis patients and skin from healthy controls (n=3), and

in the blood of psoriasis patients (n=11) and healthy controls (n=10). (b)

CD57+CD56+CD16+ NK cells in skin and the blood. (c) NKG2A expression on

CD56+CD16+ NK cells in skin and the blood. (d) NKG2C expression on

CD56+CD16+ NK cells from skin and PBMC. * = p<0.05, *** = p<0.001. Missing

values correspond to samples where number of events was too low for analysis.

index-46_1.png

35

Figure 2: Co-expression of CD57, NKG2A, and NKG2C on CD56+CD16+ NK cells

in lesional and unaffected skin of psoriasis patients and normal skin of healthy

controls. Both pie charts and bar graphs are shown. Comparisons are made between

psoriasis groups (lesional and unaffected skin) in relation to the control group (normal

skin). Significant differences (p<0.05) are represented by a (+). A significant

difference between groups was noted in CD57 and NKG2A expression alone and in

NKG2A and NKG2C co-expression. + = p<0.05.

index-47_1.png

36

Supplemental Figure 1: Lack of difference between plaque and guttate psoriasis for

the expression of markers on CD56+CD16+ NK cells. (a) CD57. (b) NKG2A. (c)

NKG2C.

37

We also examined the distribution and phenotype of CD56+CD16- NK cells. Within

the CD45+ subset, there was a trend toward an increased frequency of CD56+CD16-

NK cells in psoriasis lesional and unaffected skin compared to the skin of healthy

controls (Fig. 3a). No difference was observed in the blood. Similar to what was

observed for the CD56+CD16+ NK cell subset, CD57 expression on CD56+CD16- NK

cells was significantly reduced in lesional and non-lesional skin of patients compared

to normal skin of healthy controls (Fig. 3b). NKG2A expression was more prominent

in lesional skin compared to unaffected skin and normal skin of healthy controls,

although this trend did not meet statistical significance (Fig. 3c). No difference was

observed for NKG2C expression in the skin or blood of psoriasis patients compared to

healthy controls (Fig. 3d). No difference was observed between plaque and guttate

psoriasis (Supplemental Fig. 2).

index-49_1.png

38

Figure 3: (a) Frequency of CD56+CD16- NK cells in lesional skin (n=10) and

unaffected skin (n=10) of psoriasis patients and skin from healthy controls (n=3) and

in the blood of psoriasis patients and healthy controls. (b) CD57+CD56+CD16- NK

cells in skin and the blood. (c) NKG2A expression on CD56+CD16- NK cells in skin

and the blood. (d) NKG2C expression on CD56+CD16- NK cells from skin and

PBMC. * = p<0.05. Missing values correspond to samples where number of events

was too low for analysis.

index-50_1.png

39

Supplemental Figure 2: Lack of difference between plaque and guttate psoriasis for

the expression of markers on CD56+CD16- NK cells: (a) CD57. (b) NKG2A. (c)

NKG2C.

40

4.1.6 Discussion

Psoriasis has been largely considered a disease driven by inflammatory processes.

Much recent attention has been given to the role of Th17 and Th22 cell subsets in the

immunopathogenic process (12). Genomic studies have reinforced the importance of

aberrant inflammation by identifying relationships between inflammatory genes and

psoriasis (34, 35). Moreover, the efficacy observed with biologics targeting pro-

inflammatory elements further elucidates the contribution of cytokines such as

TNF%", IL-12, and IL-23 (7). Nevertheless, questions remain about what triggers the

onset of psoriasis that initiates the cascade of inflammation, as well as the persistence

and localization of inflammation to the involved area.

Whether NK cells are involved in the immunopathogenesis of psoriasis has been

much less studied relative to other cell types and inflammatory pathways, although

genetic and immunological analyses support a role for NK cells in psoriasis. Genetic

analyses revealed a susceptibility locus for psoriasis, the A5.1 allele of MICA, a

ligand for NKG2D (36). Others have found a psoriasis SNP between HLA-B and

MICA (37), and more recently MICA*016 was associated with psoriasis (38).

KIR2DS1 is associated with psoriasis vulgaris, and is the only activating KIR receptor

that associates to HLA-Cw6 (16-18). NK cells appear to be concentrated at the

dermal/epidermal junction (14, 28) and may have a clinically relevant role in

psoriasis, as they can function without prior sensitization and possibly act in concert

with myeloid and plasmacytoid dendritic cells (DC) to initiate plaque development.

Activated NK cells are a source of Th1 cytokines (IFN-#) and might initiate events in

psoriasis with subsequent T cell infiltration (14).

In this study, we found variations in NK cell distribution in blood and tissue of

psoriasis patients when compared to healthy subjects without psoriasis. Overall, a

trend toward higher CD56+CD16- NK cell proportions was observed in skin of

psoriasis patients compared to non-psoriasis subjects (Fig. 3a). This could be

explained as a consequence of the ongoing inflammatory process in psoriatic plaques

resulting in upregulation of integrins and chemokines that lead to increased trafficking

to the site of inflammation. CD56brightCD16- NK cells express high levels of the

41

chemokine receptors CXCR3 and CCR5, which can account for preferential homing

to the skin (28). Taken together, the skin of psoriasis patients appears to be primed for

a heightened degree of immune stimulation, at least for those in whom active

psoriasis plaques exist. The similar distribution of NK cells in lesional and non-

lesional tissue and phenotypic differences within the NK cell population suggest

underlying functional differences in NK cell maturation in psoriasis.

We, and others, recently described NK cell differentiation as a progression from a less

mature phenotype displaying high CD56 expression together with NKG2A but

lacking CD16, to a more differentiated phenotype expressing CD16 and CD57 with a

decreasing ability to replicate in vitro (27, 31, 32). Although we observed no

difference between the proportion of CD57+CD56+CD16+ NK cells in the blood of

patients and healthy controls, we observed a substantially lower percentage of

CD57+CD56+CD16+ NK cells in skin from psoriasis plaques compared to control

skin. Interestingly, non-involved skin from patients presented as an intermediate

between the two. CD57-CD56+CD16+ NK cells exhibit increased HLA-DR and

CD27 expression, as well as Ki67 and CD107a, and correspond to activated cells with

a higher turnover and degranulation ability (39). Another possible mechanism that

could explain the preferential localization of CD57- NK cells to lesions is their high

responsiveness to IL-2 stimulation, as psoriasis lesions contain high levels of IL-2

(40). Importantly, CD57-CD56+CD16+ NK cells are also more sensitive and produce

higher amounts of IFN-gamma following cytokine stimulation compared to

CD57+CD56+CD16+ NK cells. Furthermore, CD57-CD56+CD16+ NK cells may

help perpetuate the inflammation by producing high levels of IFN-gamma that

activate surrounding T cells and myeloid cells. In contrast, CD57+CD56+CD16+ NK

cells have increased KIR and granzyme B expression and correspond to more mature

cells, with lower replicative capacity, but are nonetheless functionally active (32, 39).

One previous study has analyzed CD57 expression by immunohistochemistry in

psoriasis skin, and found an increased expression of this marker in lesional skin (41).

However, that study did not identify which cells expressed CD57, and it is possible

that other cell types, such as T cells, which are more abundant in the psoriatic

inflammatory infiltrate, were CD57+.

We observed no differences in the proportion of NK cells positive for NKG2A in the

42

blood; however, there was a significant difference in the proportion of NKG2A+ cells

in lesional skin when compared to non-lesional and normal skin. NKG2A is inducible

by IL-12, which is likely to be highly expressed in the lesions, and might explain why

NKG2A is highly expressed. Alternatively, some groups have proposed that the

acquisition of functional capacities by NK cells, specifically IFN-gamma production,

is correlated with NKG2A expression on CD56+CD16+ NK cells (42). Our results are

in accordance with this observation, as lesional skin NK cells express more NKG2A.

Of interest, the expression of NKG2A was increased on circulating CD8+ T cells

from psoriasis patients, and correlated with PASI scores (43). The proportion of

NKG2C+ cells in blood showed a modest, but statistically significant difference,

whereas skin showed no such difference. This could be due to differences in CMV

serostatus because NK cells expressing NKG2C have been shown to be preferentially

expanded in CMV seropositive individuals. However, we were unable to determine

CMV serostatus on all patients in the study.

Together, these data suggest that psoriasis patients harbor a less differentiated NK cell

population in the skin. This appears to be localized to the skin itself, further