Immune cell activity during anti-TNF treatment in patients with psoriasis and psoriatic arthritis

Abstract

Psoriasis is a chronic, inflammatory skin disease characterized by a dysregulated immune response and systemic inflammation. Up to one-third of patients with psoriasis have psoriatic arthritis (PsA). Targeted treatment with antibodies neutralizing tumor necrosis factor can ameliorate both diseases. We here explored the impact of long-term infliximab treatment on the composition and activity status of circulating immune cells involved in chronic skin and joint inflammation. Immune cells were analyzed by multicolor flow cytometry. We measured markers of immune activation in peripheral blood mononuclear cell populations in 24 infliximab-treated patients with psoriasis/PsA compared to 32 healthy controls. We observed a significant decrease in the frequency of both peripheral natural killer (NK) cells and their subset CD56^dimCD16⁺ NK cells in PsA compared to healthy controls and patients with psoriasis. The latter had a strong-positive correlation with psoriasis area severity index (PASI) in these patients, while CD56^brightCD16⁻ NK cells were negatively correlated with PASI. In addition, we observed an upregulation of CD69⁺ intermediate CD14⁺CD16⁺ and CD69⁺ classical CD14⁺CD16⁻ monocytes in PsA and increased activity of CD38⁺ intermediate CD14⁺CD16⁺ monocytes in patients with psoriasis. Compared to healthy controls, psoriasis patients demonstrated shifts of the three B-cell subsets with a decrease in transitional CD27^-CD38^high B cells. Our exploratory study indicates a preserved pathophysiological process including continuous systemic inflammation despite clinical stability of the patients treated with infliximab.

Psoriasis is a chronic, inflammatory skin disease characterized by a dysregulated immune response and systemic inflammation. Up to one-third of patients with psoriasis have psoriatic arthritis. We here explored the impact of long-term infliximab treatment on the composition and activity status of circulating immune cells involved in chronic skin and joint inflammation.

Graphical Abstract

Graphical Abstract.

Introduction

Psoriasis is a chronic, immune-mediated systemic inflammatory disease manifested in the skin, in genetically predisposed individuals [1]. The most common form of the disease is plaque psoriasis, which affects about 85%-90% of psoriasis cases [2, 3]. Disease occurs bimodally, before the age of 40 (type I), and after the age of 40 (type II). Patients with psoriasis type I generally have a family history and more severe disease [4]. Red, scaly, and well-demarcated skin lesions are characterized by keratinocyte hyperproliferation and inflammatory infiltrates, steered by cytokines. However, the inflammation observed in psoriasis is not limited to the skin as indicated by higher number of circulating lymphocytes and degree of activation of peripheral blood mononuclear cells (PBMC), increased gene expression of transcription factors and cytokines involved in differentiation of T helper (Th)1, Th17, and Th22 cells as well as elevated concentrations of proinflammatory cytokines in blood [5–8]. Psoriatic arthritis (PsA) is present in almost 30% of psoriasis patients as a chronic, progressive inflammatory arthritis that leads to permanent erosions, joint destruction, and disability [9, 10]. Both disorders are associated with an increased risk of cardiovascular disease (CVD) and metabolic syndrome [11, 12].

Innovations in the treatment of moderate-to-severe psoriasis have created biopharmaceuticals that neutralize the effects of tumor necrosis factor (TNF), interleukin (IL)-12, IL-23, and IL-17A, the cytokines involved in the pathophysiological processes in psoriasis [13]. TNF is a proinflammatory mediator and one of the most important cytokines in the pathogenetic TNF/IL-23/IL-17 axis of psoriasis [13]. TNF is increased in skin lesions and its elevated level in circulation could be associated with an increase in TNF mRNA in PBMCs [14]. The consequence of TNF binding to TNF- receptors is the activation of various signal transduction pathways that regulate genes involved in cytokine transcription, as well as in the activation, differentiation, proliferation, and survival of cells involved in the development of psoriasis [15–20]. Infliximab is a chimeric monoclonal antibody which antagonizes TNF. Many patients with moderate-to-severe psoriasis and PsA have been successfully treated with infliximab in the last two decades [13, 21–23].

Recently, we could show that PBMCs from psoriasis patients treated with infliximab had higher levels of phosphorylated NF-κB, p38, ERK, and STAT3 suggesting maintenance of a systemic inflammation despite clinical remission [8].

To follow up on this for more clinical applicability, we here performed an in-depth phenotyping of PBMCs examining the expression of cell surface markers that indicate cellular activation in patients with psoriasis/PsA using multicolor flow cytometry.

Materials and methods

Patient and healthy controls

Twenty-four patients with psoriasis vulgaris with or without PsA, treated with infliximab (IFX) at the Department of Dermatology, Haukeland University Hospital, Norway, were included in this study. The patients had previously been diagnosed with moderate-to-severe psoriasis and were stable on IFX when the samples were collected. Thirty-two age, sex, and body mass index (BMI) matched healthy controls (HC) were included from the Blood bank at the Haukeland University Hospital (Table 1). The study was approved by the regional ethical committee (2014/1489 and 2014/1373). All participants provided written informed consent.

Table 1.

Characteristics of patients (n = 24) and healthy controls (n = 32) included in the study

Cohort	PsO	PsA	HC
Individuals	16	8	32
Sex (M/F)	14/2	6/2	26/6
Age, years	53.4 (16.2)	49.9 (8.6)	48.3 (13.4)
Onset of psoriasis < 40 years	14	7	NA
Psoriasis duration, years	29.4 (11.1)	24.7 (8.5)	NA
Family history	10	5	NA
BMI, kg/m2	27.9 (3.8)	28.9 (6)	26.6 (3.7)
CRP, mg/l	3.6	1.3	NA
PASI*	20.3 (10.7)	15.2 (4.4)	NA
PASI inclusion	1.9 (0.9)	1.4 (1.1)	NA
DLQI inclusion	1.1 (1.7)	0.6 (1.1)	NA
IFX** treatment (years)	4 (4.5)	5 (5.4)	NA
IFX, mg/kg	6.2 (1.4)	7.5 (1.3)	NA
IFX interval, weeks	7.2 (1.1)	7.1 (1.5)	NA
Trough level, µg/ml	12 (7.3)	10.5 (6.1)	NA
MTX, mg/week	11.9 (6.1)***	11.6 (4)****	NA
Prior biologics	10	5	NA

PsO, psoriasis patients; PsA, psoriatic arthritis; HC, healthy controls; BMI, body mass index; CRP, C-reactive protein;

*PASI before starting treatment of biologics,

**infliximab treatment before inclusion,

***15 out of 16 patients,

****8 out of 8 patients; MTX, methotrexate; NA, not applicable; values are listed as mean (SD).

Blood sampling

Peripheral blood samples were collected in lithium-heparin tubes (BD 367 526, Becton Dickinson Ltd., UK). PBMC were isolated by density gradient centrifugation with Lymphoprep (Axis-Shield Ltd., Scotland) and cryopreserved in liquid nitrogen until usage as described previously [24].

Phenotyping of PBMC

We designed a phenotyping panel of 13, fluorochrome-conjugated antibodies, and one dye [Pacific orange (PO); Supplementary Table S1] to discriminate the main mononuclear immune cell types (B cells, T cells, NK cells, NKT-like cells, and monocytes; Table 2) and analyze inflammatory markers of activation. PO dye was used as a live/dead marker. All antibodies were titrated to determine their optimal concentrations for staining.

Table 2.

Immune cell subsets defined by cell surface markers using flow cytometry

Population name	CD molecules
T cells	CD14−CD56−CD3+
T helper cells	CD14−CD56−CD3+CD4+
Naïve	CD45RO−CD27+
Central memory	CD45RO+CD27+
Effector memory	CD45RO+CD27−
Terminally differentiated	CD45RO−CD27−
T cytotoxic cells	CD14− CD56−CD3+CD8+
Naïve	CD45RO−CD27+
Central memory	CD45RO+CD27+
Effector memory	CD45RO+CD27−
Terminally differentiated	CD45RO−CD27−
NK cells	CD14−CD3−CD56+
Bright	CD56brightCD16-
Dim	CD56dimCD16+
NKT-like cells	CD14−CD3+CD56+
Monocytes	CD56−CD3−CD19−CD14+/−
Non-classical	CD14−CD16+
Intermediate	CD14+CD16+
Classical	CD14+ CD16−
B cells	CD3− CD19+
Transitional	CD27−CD38high
B naïve cells	CD27− CD38+/-
B memory cells	CD27+CD38+/−
Plasmablasts	CD27+CD38bright

For immunostaining, cryopreserved PBMC samples were rapidly thawed using a water bath set to 37 °C and washed in prewarmed X-vivo 20 containing Nuclease (1:10 000; Pierce Universal Nuclease for Cell Lysis; Thermo Fisher Scientific, MA, USA) by centrifugation at 453 × g for 5 min at 23°C. The total number of resuspended cells in prewarmed X-vivo 20 was measured with CASY cell counter (Schärfe System) and the cells were washed once in cold phosphate-buffered saline (PBS). Up to 5 million resuspended cells in PBS were stained with PO (1:1000, stock concentration: 0.5 mg/ml, Thermo Scientific) for 30 min on ice in the dark. Cells were then washed once and resuspended in FACS buffer (PBS + 0.5% BSA) with 2 µl FcR block (Miltenyi Biotec, Bergisch Gladbach, Germany) per 1 × 10⁶ cells for 5 min and stained with the antibody staining panel for 30 min in the dark at 4°C. The samples were subsequently washed once and resuspended in fixation buffer (1.6% PFA, Thermo Fisher) for 12 min at 23°C. After fixation, the samples were washed and then resuspended in FACS buffer (PBS/0.5% BSA) prior to analysis at the flow cytometer. The entire experiment was run over four consecutive days, samples were stained from one antibody master mix per day and analyzed on the same day. OneComp eBeads (eBioscience) were stained with individual antibodies and used to calculate the compensation matrix.

Samples were acquired on an LSRI Fortessa flow cytometer (BD Biosciences, San Jose, CA, USA) with BD FACSDiva™ Software (BD Biosciences) at the Bergen Flow Cytometry Core Facility, University of Bergen, Norway. The flow cytometer was equipped with 407, 488, 561, and 635 nm lasers, and emission filters for Alexa Fluor 488 (Long Pass (LP): 505, Band Pass (BP): 530/30), PerCP-Cy5.5 (LP: 685, BP: 695/40), PE (LP: —, BP: 582/15), PE-Cy7 (LP: 750, BP: 780/60), PE-TR/ECD (LP: 600, BP: 610/20), APC-Cy™7 (LP: 750, BP: 780/60), Alexa Fluor-700 (LP: 710, BP: 730/45), AF647 (LP: -, BP: 670/14), BV785 (LP: 750, BP: 780/60), BV605 (LP: 595, BP: 605/12), BV650 (LP: 635, BP: 670/30), BV421 (LP: —, BP: 450/50), BV711 (LP: 685, BP: 710/40), and PO (LP: 570, BP: 585/42). The cytometer was routinely calibrated with BD cytometer setup and tracking beads (BD Biosciences), and the application setup was used to ensure consistency of results over time. Fluorescence minus one controls were used when required for accurate gating. The positive and negative gates were defined based on biaxial plots. An average of 567 107 events in the intact cell gate was collected for each sample and the mean percentage of live cells was 96%. Flow cytometry data were analyzed in FlowJo v10.6.1 (Tree Star, Ashland, OR, USA). A representative gating strategy for a single donor is shown in Supplementary Fig. S1. To avoid the impact of differences in the count of less common populations, each sample was down-sampled to 200 000 live cell events with all analyses conducted on these down-sampled events. Down-sampling was conducted using the flowjo DownSample plugin version 3.3.1. 200 000 events were randomly sampled from cells in the “live” gate (see Supplementary Fig. S1). For UMAP visualization all samples were concatenated together. Samples belonging to HC, psoriasis, and PsA were identified, and 200 000 random events from each group were clustered using a UMAP algorithm to visualize the live cells on a 2D plain. Each cell subset was identified by manual gating, and a color key was provided. Areas of highest density are indicated in red and lowest in blue for the pseudo-color map. The heat map shows sample median fluorescence intensity (MFI) measurements of each parameter for each cell subset log 2 transformed.

Statistical analysis

Kruskal-Wallis test with an uncorrected Dunn’s test was used in the comparisons between the patients’ groups and HC in flow cytometry analyses. Differences were considered statistically significant when P ≤ 0.05. The study was of an exploratory nature and hence no correction was made for multiple comparisons. For correlation analyses between cell frequencies and disease characteristics, the strength of correlations was defined by Spearman’s rank order test. Degree of correlation was interpreted according to the recommendation of the British Journal of Medicine (https://www.bmj.com/about-bmj/resources-readers/publications/statistics-square-one/11-correlation-and-regression) with r: 0.00-0.19 (very weak), 0.20-0.39 (weak), 0.40-0.59 (moderate), 0.60-0.79 (strong), and 0.80-1.00 (very strong). Comparisons between groups and the creation of violin and correlation plots were done using GraphPad Prism (Version 9.1.1 (225)). Volcano plots and heatmaps were generated in Microsoft Excel.

Results

Twenty-four patients (Table 1), of whom 16 with psoriasis (PsO) and 8 with psoriasis and PsA on continuous IFX treatment, were compared to 32 age, sex, and BMI matched HC. A schematic overview of the experimental design and analysis workflow is given in Supplementary Fig. S2.

PsO and PsA groups display skewed immune cell frequencies

Visualization of live PBMC through UMAP clustering (Fig. 1A) of pooled events indicated shifts in the immune cell profiles between the 3 groups. These differences were most apparent in the NK-cell populations as identified by manual gating, where the UMAP visualization indicated a decrease of NK cells in PsA. Statistical comparisons between the groups (Kruskal-Wallis test with an uncorrected Dunn’s test; Fig. 1B, C; Supplementary Tables 2 and 3) as displayed in volcano and violin plots showed this was indeed the case with a decreased frequency in both NK cells (median 7.0) and the CD56^dimCD16⁺ NK-cell subset (median 4.3) in these patients compared to both HC (NK cells P = 0.0378; CD56^dimCD16⁺ NK cells P = 0.0085) and PsO (NK cells P = 0.0227; CD56^dimCD16⁺ NK cells P = 0.0112). This decrease of the CD56^dimCD16⁺ NK-cell subset was further highlighted by examining frequencies within NK cells, with a decrease of the CD56^dimCD16⁺ NK-cell subset corresponding with an increase in the frequency of the CD56^brightCD16^- NK-cell subset (median 10.8) when comparing HC to PsA (P = 0.0333) (Fig. 1C). Interestingly, shifts of the three B-cell subsets (naïve CD27⁻ CD38^+/−, transitional CD27^-CD38^high and memory CD27⁺CD38^+/− B cells) were observed, with an increase in memory CD27⁺CD38^+/− B cells (median 38.0) and a decrease in transitional CD27⁻CD38^high B cells (median 0.08) and naïve CD27^- CD38^+/− B cells (median 48.1) in PsO compared to HC (memory CD27⁺CD38^+/- B cells P = 0.0417, transitional CD27^-CD38^high B cells P = 0.0450, naïve CD27^- CD38^+/- B cells P = 0.0235), while no shifts were apparent in the PsA patients (Fig. 1C).

Figure 1:

patient groups and healthy controls display differences in frequencies of PBMC subsets. (A) Cell subsets. Cells were gated by scatter properties (SSCA vs FSCA), then single cells (FSCH vs FSCA; SSCH vs SSCA), followed by live cells (SSCA, PO⁻). Each sample was then sampled down to 200 000 live cell events. All samples were concatenated together. Samples belonging to healthy controls (HC), psoriasis (PsO), and psoriatic arthritis (PsA) patients were than identified, and 200 000 events from each group were clustered using a UMAP algorithm using all parameters except PO and scatter properties to visualize the live cells on a 2D plain. Parental populations (CD4⁺ T cells, CD8⁺ T cells, monocytes, B cells, and NK cells) are indicated. Each cell subset identified by manual gating (see Supplementar Fig. S1) is indicated by color. Pseudo-color density plots indicate distribution of cells with areas of highest density indicated in red and lowest in blue. Heatmap of log transformed median marker intensity normalized to a range of 0 to 1 of the 13 markers for each of the 18 identified populations indicated on the righthand side of the figure. (B) Results of comparisons of cell frequencies normalized to total live cells, between the 3 groups are summarized in volcano plots and given in the top row, and results of comparisons of cell frequencies normalized to their respective parent (B, monocytes, CD4⁺ T cells, CD8⁺ T cells, and NK cells), between the 3 groups are given in the bottom row. Significant values (P ≤ 0.05) are indicated as red circles, and cell designations are given for significant values and those approaching significant. (C) Comparisons between the groups displayed in violin plots show frequency in NK- and B-cell populations and subsets

Activation marker expression in immune cell differ significantly between PsO and PsA groups

Visualization of live PBMC through UMAP clustering (Fig. 2A) of pooled events indicated no strong differences in expression of activation markers (CD38, CD69, HLA-DR, and CD107a) between the 3 groups except for CD69 on monocytes where a notable increase in PsA patients was observed. Closer examination with a statistical comparison (Fig. 2B, C) between the groups confirmed the increased CD69 expression on monocytes in PsA where CD69 was increased in both intermediate CD14⁺CD16⁺ (median 652.5) and classical CD14⁺CD16^- monocytes (median 730.5) compared to both HC (intermediate CD14⁺CD16⁺ monocytes P = 0.0154; classical CD14⁺CD16^- monocytes P = 0.0191) and PsO (intermediate CD14⁺CD16⁺ monocytes P = 0.0130; classical CD14⁺CD16⁻ monocytes P = 0.0071; Supplementary Table S4). A skewed CD38 expression level (Supplementary Table S5) in multiple immune cell subsets was the most notable feature in PsO, with increased expression in the CD4⁺ effector memory CD45RO⁺CD27⁻ (median 10 452), and terminally differentiated CD45RO⁻CD27⁻ (median 4182) T-cell subsets, as well as in intermediate CD14⁺CD16⁺ monocytes (median 12 496) and NKT-like cells (median 19 119) compared to HC (CD4⁺ T_EMP = 0.0093; CD4⁺ T_TDP = 0.0045; intermediate CD14⁺CD16⁺ monocytes P = 0.0206; NKT-like cells P = 0.0101). Interestingly, both patient groups displayed reductions of CD38 on memory CD27⁺CD38^+/− B cells (PsO median 8.0; PsA median 8.4) compared to HC (PsO P = 0.0003; PsA P = 0.0164). Although differences were small, a consistent decrease of CD107a expression on both T (CD8⁺ T_CM median 484.0; CD8⁺ T_TD median 468.0) and NK-cell subsets (CD56^brightCD16^- NK cells median 315.0; CD56^dimCD16⁺ NK cells median 339.5) was notable in PsA (CD8⁺ T_CMP = 0.0009; CD8⁺ T_TDP = 0.0226; CD56^brightCD16^- NK cells P = 0.0148; CD56^dimCD16⁺ NK cells P = 0.0006), interestingly this decrease was limited to CD56^dimCD16⁺ NK-cell (median 364.0) and CD56^brightCD16⁻ NK-cell (median 302.0) subsets in PsO (to CD56^dimCD16⁺ NK cell P = 0.0031; CD56^brightCD16⁻ NK cell P = 0.0042; Supplementary. Table S6).

Figure 2:

expression of activation markers CD38, CD69, CD107a, and HLA-DR on immune cells. Degree of expression was examined using the 95 percentiles of the MFI. T and NK subsets, and NKT cells were examined for CD38, CD69, CD107a, and HLA-DR. Monocytes were examined for CD38, CD69 and HLA-DR. B-cell subsets were examined for CD38 and HLA-DR. (A) MFI measurements of CD38, CD69, CD107a, and HLA-DR are shown using a UMAP clustering for each sample subtype. (B) Results of comparison between each sample group of 95 percentiles of CD38, CD69, CD107a, and HLA-DR expression in different cell subsets. Cell designations are given for significant values (P ≤ 0.05), tags are colored based on marker. (C) Expression of activation markers CD69, CD107a, and CD38 between the 3 groups displayed in violin plots

Correlation analyses

PsA patients showed strong negative correlation between severity of psoriasis, as assessed by PASI, and NK-cell subset CD56^brightCD16⁻ cells (r = −0.8810; P = 0.0072; Fig. 3A) and strong positive correlation between PASI and NK-cell subset CD56^dimCD16⁺ cells (r = 0.7381; P = 0.0458; Fig. 3B). Moreover, very strong positive correlation was observed between PASI and CD8⁺T_CM cells (r = 1000, P < 0.0001; Fig. 3C). The absence of correlation between investigated cell frequencies and PASI was observed in patients with PsO. However, in PsO patients, a strong-to-moderate negative correlation of peripheral B-cell frequencies and patient age at inclusion was observed (B cells (r = −0.7535; P = 0.0011), naïve B cells (r = −0.5107; P = 0.0452) memory B cells (r = −0.6853; P = 0.0044; Supplementary Fig. S3A–C)). In patients with PsA, this negative correlation was only observed for memory B cells (r = −0.8783; P = 0.0068; Supplementary Fig. S3D). In contrast, the frequency of B cells, especially naïve B cells, showed a strong and very strong positive correlation (B cells (r = 0.7785, P = 0.0287), naive B cells (r = 0.8264, P = 0.0158)) with disease duration in PsA patients, which was not found in patients with PsO (Supplementary Fig. S3E, F). No correlation was found between cell populations and BMI of patients, nor treatment duration.

Figure 3:

correlation of cell subsets with PASI in patients with PsA. (A) Very strong negative correlation of CD56^brightCD16^- NK cells with PASI. (B) Strong positive correlation of CD56^dimCD16⁺ NK cells with PASI. (C) Very strong positive correlation of CD8⁺ T_CM cells with PASI

Discussion

Motivated by a lack of studies investigating the activity state of peripheral blood immune cells in patients with psoriasis on continuous infliximab treatment, we conducted this analysis to trace PsO/PsA-specific immune profiles and/or indicators.

In concordance with previous findings that psoriasis patients have deviant immune cell frequencies which normalize with TNF treatment, no differences in cell frequencies of main populations except for NK cells were observed [24]. The decrease of circulatory NK cells relative to total live PBMC in PsA compared to HC and PsO could indicate intensified disease activity due to the ongoing NK-cell recruitment from the circulation to synovium, as suggested in previous studies [25]. Since our patients were on stable IFX treatment with low disease activity, we can only speculate that the decrease in peripheral NK, i.e. CD56^dimCD16⁺ NK cells in PsA could be a result of chronic systemic inflammation with impaired cell survival [26]. In addition, the state of NK-cell activity through the expression of the investigated surface receptors did not differ from HC, except for CD107a, which was significantly reduced in both NK-cell subpopulations of patient groups, thus indicating their reduced cytotoxicity [27].

Numerous studies have observed a contributing role of B cells in the pathophysiology of psoriasis [28]. Opposing results have shown variations in quantity and quality of peripheral B cells and their subsets depending on disease activity and applied treatment [29–31]. In our patients, the disturbed ratio of B-cell subsets with the shift from naïve CD27⁻ CD38^+/− to memory CD27⁺CD38^+/− B cells and significant reduction of transitional CD27⁻CD38^high B cells was in concordance with a previous study [32]. In humans, immature transitional B cells, through IL-10 secretion, display regulatory functions [33]. Transitional B cells suppress autoreactive CD4⁺ T-cell proliferation, CD8⁺ T cells activation, production of proinflammatory cytokines, differentiation of CD4⁺ T cells into Th1 and Th17 and contribute to the transformation of effector CD4⁺ T cells into functional Tregs [34–36]. Recent studies have shown that transitional B cells are decreased and functionally impaired in both, PsA and psoriasis, in the circulation and skin [37–39]. In our PsO patients, a decrease in transitional B cells could indicate their reduced suppressive capacity contributing to the dysfunctional state of Tregs, even though patients were on stable IFX treatment. Unlike PsO, patients with PsA showed complete restoration of B-cell frequencies as noted in previous research [25]. Although memory CD27⁺CD38^+/- B cells were increased in PsO, their activity in both patient groups was reduced through the decrease of CD38 which is fundamental in BCR activation [40]. Downregulation of CD38 was probably IL-4 induced as has been described in IFX treated patients with rheumatoid arthritis [41–43].

Recent studies have shown positive correlation between B-cell populations and PASI in patients with PsO and PsA, and negative correlation with the PASI of pustular psoriasis [44]. No correlation between severity of psoriasis and NK cells has been observed in published research [45].

Different studies have observed that psoriasis is associated with a higher prevalence of cardiovascular risk factors, such as obesity, high blood pressure, diabetes mellitus, dyslipidemia, and metabolic syndrome [44–46]. Pathophysiology of psoriasis and atherosclerosis has a common denominator, such as monocytes [47]. Elevated levels of circulatory intermediate monocytes and reduced levels of classical monocytes, both with proatherogenic properties, have been observed in psoriasis [24, 48]. In our study, PsO had higher frequencies of intermediate CD14⁺CD16⁺ monocytes. It has been shown that these cells could be involved in creation of monocyte-platelet aggregates (MPA) with consequent increased adhesion to vascular endothelium and trans-endothelial migration [49]. As a marker of a platelet activation, MPA is an indicator of coronary artery disease that may contribute to atherosclerosis progression [50]. Moreover, patients with psoriasis have a higher risk of death due to cardiovascular disease (CVD) caused by an increase in MPA, and intermediate monocytes that correlate with severity of the disease [48, 51]. Conflicting data exists regarding the effect of infliximab treatment to the risk of CVD in patients with severe psoriasis [52, 53]. Compared to conventional systemic therapies, or phototherapy, treatment with TNF inhibitors reduces the risk of CVD in patients with severe psoriasis [52–54]. All but one patient were treated with IFX in combination with methotrexate, and methotrexate treatment alone was also shown to result in lower CVD event rates [54], even after long-term follow-up [55]. However, there is a lack of evidence that can equalize reduced CVD risk by TNF inhibitors in psoriasis to CVD risk in a general healthy population. In our study, an elevated level of activated intermediate CD14⁺CD16⁺ monocytes through increased expression of CD69 in PsA and CD38 in PsO supports the hypothesis of ongoing systemic inflammation in both patient groups, even after resolution of skin lesions. As systemic inflammation promotes CVD, this result raises the question if these patients have a persisting increased risk of CVD due to the activity of different surface molecules and poor balance between pro- and anti-atherogenic monocytes [54, 56].

CD107a is a lysosomal membrane protein present on lymphocytes which is capacitated to release perforin containing granules and can therefore be used as a marker for cytotoxicity [27]. Psoriasis patients with the most pronounced Psoriasis Area Severity Index (PASI), Body Surface Area (BSA), and Dermatological Life Quality Index (DLQI) scores have higher incidence of CD107a⁺ cells in epidermis compared to dermis, without the precise identification of cell types [57]. Our analysis of PBMC showed reduced cytotoxic activity in innate and adaptive immune cells including above mentioned CD56^brightCD16⁻ NK cells and CD56^dimCD16⁺ NK cells in both patients groups as well as subpopulations of CD8⁺ T cells (T_CM, T_TD) that were restricted to PsA as a confirmation of the favorable effect of IFX treatment [27].

Within immune cells, CD38 is expressed constitutively on plasmablasts and plasma cells, and after stimulation on macrophages, NK cells, monocytes, NKT-like cells, T cells, DCs, and innate lymphoid cells [58]. Our finding of the increased expression of CD38 on CD4⁺ T_EM in PsO patients could point to their greater ability to traffic through peripheral organs and blood [59]. In addition, the role of elevated CD38 on CD4⁺ T_TD cells in PsO is not known although we can speculate that these cells have an active profile with reduced proliferation capacity but improved potential to produce cytokines as it has been observed in studies conducted both, in mice and humans [60, 61]. This raises the question of whether antigens/autoantigens in psoriasis trigger terminal effector differentiation and T_TD/T_EMRA cell formation, which may be related to antigen loading or persistence. It would be interesting to simultaneously analyze expression of CD38 on CD4⁺ T_TD cells in lesional skin and circulation and explore their capacity to produce cytokines.

CD3⁺CD56⁺ NKT-like cells are involved in the pathogenesis of psoriasis [29]. Their numbers in skin inflammatory infiltrate decrease with the treatment [62]. In untreated patients with psoriasis, levels of circulating NKT-like cells are stable [63]. However, recent findings in patients with RA have shown that CD38⁺ NKT-like cells co-cultured with T cells stimulate the differentiation of CD4⁺ T cells into Tregs and reduce the Th17 level in synovial fluid [64]. In our study, higher expression of CD38 on NKT-like cells in PsO but not in PsA could indicate a beneficial effect of IFX treatment, but it remains unknown whether these cells reach lesional skin and restore impaired immune tolerance, and whether they are sufficient for the normal functioning of Tregs, which are already disturbed in psoriasis.

A limitation of the study is the small sample size that consequently affects the statistical analysis, lack of markers for analysis of CD4+ T-cell subpopulations (Treg, Th1, Th2, and Th17), as well as lack of skin biopsy specimens of psoriatic lesions and synovial tissue of PsA, and lack of joint disease activity score for PsA patients. All patients except one patient with PsO were treated with infliximab and methotrexate simultaneously. It has been shown that methotrexate helps in the restoration of the immune balance by decreasing Th1 and Th17 cells and increasing Th2 and Tregs in circulation, thus resulting in a significant reduction of disease severity [65]. However, it is not known whether this effect can be seen with IFX alone. The functionality of the activated cells needs also to be evaluated in future studies. We also cannot exclude that the effects seen in our study are specific for IFX. Future studies will need to confirm if other TNF inhibitors have a similar effect. Moreover, as for all studies using cryopreserved PBMC, selective loss of distinct cell populations due to cryopreservation and storage time cannot be excluded. However, PBMC are considered to be rather stable during long-term storage [66, 67], even though B cells and myeloid cells tend to be more affected by cryopreservation compared to freshly isolated cells [67].

Conclusions

Our exploratory study on the effect of stable biological treatment showed that patients with PsO or/and PsA still retain imprint from the disease and have phenotypic peculiarity. Beyond clinical stability, what they still have in common is a systemic inflammation that persists despite stable IFX treatment.

Supplementary data

Supplementary data is available at Clinical and Experimental Immunology online.

Glossary

Abbreviations:

BMI

body mass index

BSA

body surface Area

CVD

cardiovascular disease

DLQI

dermatological life quality index

HC

healthy controls

IFX

infliximab

IL

interleukin

NK

natural killer

MFI

median fluorescence intensity

MPA

monocyte-platelet aggregates

PASI

psoriasis area severity index

PBMC

peripheral blood mononuclear cells

PO

pacific orange

PsA

psoriatic arthritis

PsO

psoriasis

Th

T helper

TNF

tumor necrosis factor

Ethical approval

The studies involving humans were approved by Western Norway ethical committee (approval 2017/938). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Conflict of interest

The authors declare no conflict of interest.

Funding

This study was supported by the Faculty of Medicine, University of Bergen, the Western Norway Regional Health Authority (grant number 912065), Meltzer foundation, and Broegelmann foundation.

Data availability

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.