A Western diet, but not high fat and low sugar diet, predisposes mice to enhanced susceptibility to imiquimod-induced psoriasiform dermatitis

To the Editor

Psoriasis is a disease with systemic inflammation and accompanied by multiple co-morbidities, including metabolic syndrome and obesity (Hwang et al., 2017, Tsai et al., 2017, Yu et al., 2017). Obesity is often observed in patients with psoriasis and may precede development of psoriasis (Debbaneh et al., 2014). Western diet (WD) plays a crucial role in the development of obesity in the Western countries and is characterized by elevated amounts of fat and sugars, especially simple sugars such as sucrose (Jena et al., 2017). Recent research has revealed that, in a murine model, WD triggers systemic inflammation via the NLRP3 inflammasome and subsequent production of pro-inflammatory cytokines such as interleukin (IL)-1β (Christ et al., 2018). On the other hands, a high fat diet (HFD) without excessive sugars is often used to establish murine models of obesity. Herein, we established two models of obesity by feeding mice with WD or HFD. The WD is not only rich in fat but also contains high content of sucrose, replicating the type of diet in the Western world that contributes to obesity. The HFD obtains 60% of calorie from fat and has a sucrose content that is similar to regular mouse chow. We compared these two kinds of diet-induced obese mice with lean mice on control diets in terms of susceptibility to imiquimod (IMQ)-induced psoriasiform dermatitis (PsD). Animal protocols were approved by the Institutional Animal Care and Use Committee at the University of California, Davis. Of interest, mice on HFD gained more weight than mice on WD by week 7 of feeding and sustained this increased body mass over the course of the experiment (Figure 1a, b). The measurement of ear thickness change following exposure to topical IMQ (Cochez et al., 2017, Yu et al., 2018) or intradermal IL-23 (Mabuchi et al., 2011) is a standard method to assess the extent of PsD. Although mice fed with HFD had the greatest weight gain, mice fed with WD for 12 or 16 weeks had significantly more ear thickness change than mice fed with control diet (CD), HFD, and low fat diet (LFD) after a 5-day IMQ treatment course (P<0.0001) (Figure 1c, d). There was no difference in ear thickness change between HFD- and LFD-fed mice on day 5. Thus, mice on WD, but not HFD, had enhanced susceptibility to IMQ-induced PsD 3–4 months after sustained feeding despite greater weight gain in HFD-fed mice. WD-fed mice showed consistently greater epidermal hyperplasia than the other three dietary groups (Figure 1e, f). Because neutrophils are a characteristic feature of psoriatic lesions and may play a key pathogenic role in psoriasis, we quantified neutrophil abscess formation as well as neutrophil chemoattractants in the four dietary groups. While there was no difference in Cxcl1 expression, WD-fed mice had higher expression of Cxcl2 than HFD-fed mice after 5-day IMQ treatment (Figure 1g, h). WD-fed mice had the highest expression level of neutrophil marker Ly6g mRNA and the highest density of Munro microabscess (Figure 1i, j). Consistent with RT-PCR findings, more Gr-1(+) cells were observed in IMQ-treated WD-fed mice than other groups (Figure 1k, l). Expression levels of Ly6c mRNA were not different among four dietary groups, which indicates higher expression of Gr-1 in WD-fed mice results from higher expression of Ly6g but not Ly6c (Supplementary Figure 1). Therefore, compared to the other three dietary groups, WD-fed mice had increased epidermal hyperplasia in response to IMQ as well enhanced expression of a neutrophil chemoattractant, which may explain the histologic appearance of IMQ-treated skin in the WD-fed mice. Our previous research has indicated CCR6+ Th17 cells and γδ T cells secretes IL-17A in murine psoriasiform models (Mabuchi et al., 2013). Because IL-17A is a key mediator of neutrophilic inflammatory states, we postulated that IL-17A may help explain the enhanced neutrophilic infiltration observed in WD-fed mice after IMQ treatment. As anticipated, IMQ induced higher gene expression of Il-17a in all four dietary groups (Figure 2a). To determine if the different diets could alter baseline expression of inflammatory cytokines in skin, we next assessed their expression in vehicle-treated ears, which reflect expression of these cytokines in the absence of IMQ stimulation. Strikingly, WD induced a 45-fold increase of Il-17a compared with CD while HFD induced only 20-fold increase (WD vs. HFD, P=0.050). Gene expression levels of IL-22 and IL-23, both key psoriasis-related cytokines, were highly elevated in IMQ-treated skin of the various dietary groups as expected, but were not different in the vehicle-treated ears of HFD and WD mice compared to control counterparts (Figure 2b, c). Thus, WD mice show a specific trend toward elevated Il-17a, but not Il-22/23, in the skin, potentially explaining the enhanced reactivity to IMQ. Of note, WD-fed mice had the highest expression level of Nlrp3 and pro-inflammatory cytokine Il-1b genes among four dietary groups after a 5-day course of IMQ. As compared with HFD-fed mice, WD-fed mice had higher mRNA levels of Nlrp3 (P=0.006; Figure 2d) and Il-1b (P=0.020, Figure 2e) in IMQ-treated ears while the expression of Tnf-a showed no difference (Figure 2f).

Figure 1. WD-fed mice have enhanced susceptibility to psoriasiform dermatitis and higher expression levels of neutrophil markers and developed more Munro microabscesses in response to IMQ treatment.

(a) Mice fed with high fat diet gained more body weight than those fed with Western diet, low fat diet, or control diet. (b) Photographs of mice in four different diet groups after feeding for 16 weeks. (c) At week 12, mice fed with WD had significantly higher ear thickness change than mice fed with HFD on day 4 and day 5 of the IMQ course. No difference in ear thickness change was observed between HFD mice and LFD mice on day 5. (d) At week 16, WD-fed mice had more prominent ear thickness change than HFD-fed mice from day 3 to day 5 of the IMQ course. There was no difference in ear thickness change between HFD and LFD mice on day 5. (e) Epidermal thickness after 5-day IMQ or vehicle treatment as measured in microscopic fields. Vehicles-treated ears of HFD and WD mice were thicker than LFD and CD mice, respectively. IMQ-treated ears of WD mice were thicker than those of HFD mice. Vehicles-treated ears of WD mice were also thicker than those of HFD mice, although the difference did not reach statistical significance. (f) Representative photos of H&E staining of ears after 5-day IMQ treatment in four groups (bar, 50 μm). More cell infiltrates and parakeratosis were observed in WD mice. A wild-type (WT) mouse ear treated with 5-day vehicle cream was shown for comparison. (g, h) While there was no difference in Cxcl1 expression between WD-fed and HFD-fed mice, WD-fed mice had higher expression of Cxcl2 after 5-day IMQ treatment. (i) WD-fed mice had the highest mRNA expression of the neutrophil marker Ly6g. (j) Consistent with the highest expression of Ly6g, there were more Munro microabscesses observed in WD-fed mice. (k) More Gr-1 positive cells were observed in WD-IMQ group than other three IMQ groups. (l) Representative photos of Gr-1 immunohistochemistry. Black scale bar: 20 μm. HFD, high fat diet. LFD, low fat diet. WD, Western diet. CD, control diet. IMQ, imiquimod. * P<0.05. ** P<0.01. *** P<0.001. **** P<0.0001

Figure 2. WD-fed mice have the highest baseline mRNA expression of Il-17a and, after IMQ 5-day course, have the highest mRNA expressions levels of Nlrp3 and Il-1b among the four dietary groups.

(a) IMQ treatment induced higher expression levels of Il-17a in all four dietary groups. Then we examined baseline expression of Il-17a in vehicle-treated ears. WD induced a 45-fold increase of Il-17a compared with CD while HFD induced only 20-fold increase (WD vs. HFD, P=0.050). (b, c) WD and HFD did not increase expression levels of Il-23 and Il-22 in both IMQ- and vehicle-treated ears. (d, e) After IMQ 5-day course, WD-fed mice had higher expression of Nlrp3 and Il-1b than HFD-fed mice. (f) There was no difference in the expression of Tnf-a among four dietary groups. IMQ, imiquimod. NS, no statistical significance. * P<0.05. ** P<0.01.

The mechanisms how sugar exacerbates psoriasis are likely multifactorial. Reports suggest that WD triggers systemic inflammation via the NLRP3/IL-1β activation (Christ et al., 2018), potentially with subsequent induction of neutrophil chemoattractants such as CXCL1 (Amaral et al., 2012). Another possibility is that sugar intake could change gut and skin microbiota and modify cutaneous immune microenvironment to potentiate skin to psoriasiform dermatitis. For example, one study shows that Western diet induces a shift in gut microbiota composition and enhances susceptibility to E. coli infection and intestinal inflammation (Agus et al., 2016). Dietary components not only change gut microbiota but influence cutaneous inflammation via gut-skin axis (Maguire and Maguire, 2017, O’Neill et al., 2016). It is intriguing that CXCL2, but not CXCL1, is elevated in WD-fed mice. Although CXCL1 and CXCL2 do regulate NLRP3 inflammasome activation via the same G protein-coupled receptor CXCR2 (Boro and Balaji, 2017), literature has indicated that CXCL1 and CXCL2 do not necessarily contribute equally (Ritzman et al., 2010), suggesting that non-parallel induction of one or the other may occur in some situation.

The fact that HFD-fed mice have more body weight gain but less inflammation in response to IMQ than WD-fed mice strongly suggests obesity alone is not sufficient to promote PsD in the skin. Intriguingly, a recent study indicated that a high dietary fat diet exacerbates psoriatic skin inflammation independent of obesity (Herbert et al., 2018). Another study, using different chow recipes, also concluded that high dietary fats exacerbates IMQ-induced psoriasis-like dermatitis in mice (Higashi et al., 2018). Despite being termed “high fat diets” by their authors, these two studies used diets which were not only rich in fat but also contained a high content of simple sugars, bearing considerable similarity to the WD used in our studies (Supplementary Table 1). Based on our findings and their results, we propose that dietary content, beyond just the ability to induce obesity, plays a crucial role in the predisposition to psoriasiform inflammation. In conclusion, it is a WD, but not a HFD alone, that predisposes mice to enhanced susceptibility to IMQ-induced PsD. Although we cannot say that higher than standard levels of fat in the WD are dispensible for increased PsD susceptibility, our data do suggest that sucrose/sugar content of the WD is a critical factor for the enhanced susceptibility to IMQ-mediated PsD.

Abbreviations:

ANOVA

analysis of variance

CD

control diet

HFD

high fat diet

IL

interleukin

IMQ

imiquimod

LFD

low fat diet

PsD

psoriasiform dermatitis

WD

Western diet