IL-12/IL-23 neutralization is ineffective for alopecia areata in mice and humans

Capsule Summary:

Ustekinumab is currently being explored for the treatment of alopecia areata. Here, our work in a murine model and in human patients indicates that neutralization of the IL-12 and IL-23 pathways does not ameliorate disease.

To the Editor:

Alopecia areata (AA) is a common autoimmune disease resulting in non-scarring hair loss. Recent work involving genome wide-association studies¹, whole genome gene expression assessments², and mechanistic studies in animal models³ have revealed the role of interferon (IFN)-γ and common γ-chain cytokines, including interleukin (IL)-15 and IL-2, in the pathogenesis of AA, as well as the central role of NKG2D-expressing CD8 T cells as pathogenic disease mediators. These data have spurred investigations into novel treatments and have uncovered the potential for repurposing agents already approved for human use to treat AA, including Janus kinase (JAK) inhibitors^4,5.

It was recently reported that expression of the genes encoding the shared p40 subunit of IL-12 and IL-23 and the IL-23-specific p19 subunit were increased in skin biopsy samples from AA lesions when compared with that from nonlesional scalp⁶ and that effective treatment of AA lesions with intralesional administration of triamcinolone led to the reduction of expression of p40 but not p19⁷. In addition, AA patients treated with ustekinumab, a human monoclonal antibody that binds the shared p40 subunit of IL-12 and IL-23 and antagonizes their activities, demonstrated hair regrowth⁸. These data and subsequent case series^9,E1 suggest that ustekinumab may be a potential therapy for AA. Interestingly, despite data indicating that IFN-γ, a Th1 cytokine, is critical for AA pathogenesis³, it was reported that the p35 subunit specific for the Th1-inducing IL-12 cytokine was not elevated in lesional AA samples compared with nonlesional control samples. These data hint that the IL-23-altering effects of ustekinumab may be responsible for its efficacy in AA.

To further investigate the role of IL-23 and IL-12 in the pathogenesis of AA and the potential for targeting the shared p40 subunit for treatment purposes, we first assessed the levels of the p40 subunit in the C3H/HeJ mouse model of AA. Levels of p40 in the skin of AA mice were found to be elevated when compared with unaffected controls (Figure 1A). Similarly, we found that levels of the IL-23-specific p19 subunit and the intact IL-12 heterodimer were elevated in affected skin of AA mice compared to unaffected controls (Figure 1A). Based on these data, we next determined whether neutralization of the shared p40 subunit or the IL-23-specific p19 subunit were effective at suppressing the development of AA. Using a previously described method of AA induction involving adoptive transfer of in vitro expanded lymph node cells from a previously affected C3H/HeJ mouse^E2, C3H/HeJ lymph node cell recipients were treated twice weekly with antibodies specific for the IL-12/IL-23 p40 subunit, the IL-23 p19 subunit, or control antibody and assessed for the development of hair loss. Treatment with antibodies specific for IL-12/23 p40 or IL-23 p19 did not affect the development of hair loss in C3H/HeJ mice compared with mice treated with IgG control (Figure 1B). Mice treated with these antibodies did not exhibit significant differences in the frequency of AA development or the speed with which disease developed (Figure 1C). The mice that developed AA likewise exhibited the emergence of NKG2D-expressing CD8 T cells in skin draining lymph nodes (Figure 1D), supporting that neither IL-12 nor IL-23 are required for the development of these pathogenic disease effectors. Similar data were obtained in antibody treated mice induced by AA skin graft transfer ^E3 (data not shown). Finally, we confirmed that antibody treatment effectively neutralized the appropriate cytokines; treatment with a neutralizing antibody to the shared p40 subunit reduced levels of circulating p40 as well as intact IL-12, and treatment with antibodies specific for the IL-23-specific p19 subunit reduced levels of circulating IL-23p19 (Figure 1E).

Figure 1. Targeting the shared p40 subunit does not prevent the development of murine AA.

Levels of IL-12/IL-23 p40 subunit and IL-23 specific p19 subunit compared were assessed by ELISA in AA mice or unaffected control mice (A). Expanded AA LN cell-induced C3H/HeJ mice were treated with α-p40, α-p19 or IgG control antibodies twice weekly starting at the time of transfer. Representative photographs of mice from each group are shown (B). Mice were observed for the development of AA over time (C). Skin-draining lymph nodes were harvested and stained for NKG2D and CD8 (D). Serum levels of the p40, p19, or the intact IL-12 heterodimer are shown (E). Data represents mean with SD. A and B was analyzed with unpaired t test and E with ANOVA followed by Kruskal-Wallis test (* p<0.05; ** p<0.01; *** p<0.001).

A retrospective chart review was subsequently conducted to determine the potential efficacy of targeting the IL-12/IL-23 p40 subunit for treatment purposes in human AA. Four patients with AA with complete or near-complete scalp hair loss were treated with ustekinumab (Table 1). All patients had previously tried and failed multiple therapies, although some patients did demonstrate incomplete responses (as indicated in Table 1). There was no hair regrowth with ustekinumab treatment in any of the patients, all of whom had relatively short duration of complete or near-complete scalp hair loss and were treated with high-dose ustekinumab. Interestingly, after discontinuing ustekinumab, patient 1 experienced complete hair regrowth (and improvement in her atopic dermatitis) after starting tofacitinib 5 mg twice daily.

Table 1: Baseline characteristics of patients with alopecia areata and treatment history and outcomes.

AA: alopecia areata; AU: alopecia universalis; AD: atopic dermatitis; DMT2: diabetes mellitus type 2; ILTAC: intralesional triamcinolone; PUVA: psoralen and ultraviolet A; TCI: topical calcineurin inhibitors; TCS: topical corticosteroids

Characteristic	Patient 1	Patient 2	Patient 3	Patient 4
Age (yr)	8	16	14	44
Gender	F	M	M	F
Duration of AA (yr)	1	4.5	5	33
Duration of current episode of complete or near-complete scalp hair loss (yr)	1	3	4	4
Pre-treatment SALT score (%)	100	100	100	100
Post-treatment SALT score (%)*	100	100	100	100
Percent change in SALT score (%)	0	0	0	0
Ustekinumab dose and frequency (total No. injections)	45 mg at 0, 60 mg at 1, 90 mg at 3 and 5 months (4)	90 mg every 8 weeks for 30 months (15)	90 mg at 0, 1, and 3 months (3)	90 mg at 0, 1 and 3 months (3)
Autoimmune comorbidities	AD	AD, hay fever, asthma, Crohn’s	None	Hypothyroidism
Family history of autoimmune disease	None	None	Hypothyroidism	Thyroid disease
Previous treatments	TCS, ILTAC, cyclosporine A	TCS, ILTAC	TCS, TCI, ILTAC, prednisone, squaric acid immunotherapy, tofacitinib	TCS, ILTAC, methotrexate, ruxolitinib, tofacitinib
Efficacy of previous treatments	None†	None	Partial regrowth of scalp (tofacitinib) and eyebrows (ILTAC)	Regrowth of eyelashes (ruxolitinib)

^*Patients were evaluated in all cases within 8 weeks of the last administered dose

^†After ustekinumab treatment, patient experienced complete scalp hair regrowth (and remission of her atopic dermatitis) taking tofacitinib 5 mg twice daily

The potential role of IL-23 in the pathomechanism of AA has not been clearly defined. IL-23 plays a role in the differentiation of naive CD4 T cells into Th17 cells and is critical to their maintenance and survival. Th17 cells are thought to be important for coordinating the defense against extracellular bacteria, have been implicated in specific autoimmune diseases including psoriasis, inflammatory bowel disease and multiple sclerosis, and participate in the recruitment of neutrophils to sites of inflammation, as seen in psoriatic microabscesses but not in skin biopsy specimens of AA. Although one study indicated that IL-23 and IL-23-associated genes exhibit increased expression in skin biopsy specimens from patients with AA⁶, this was not consistent among all studies². Additionally, treatment of patch-type AA lesions with intralesional glucocorticoids, resulting in partial to excellent hair regrowth, showed no difference in IL-23 p19 expression before and after successful treatment⁷.

Regarding IL-12, a potential role for this cytokine in the development of AA seems more plausible, and previous reports associating decreased IL-12/23 p40 subunit gene expression following intralesional triamcinolone treatment may be interpreted as suggesting IL-12 involvement in AA⁷. IL-12 may augment CD8 T cell responses and is important for the differentiation of naive CD4 T cells into Th1 effectors; Th1 effectors produce IFN-γ, a cytokine that is critical to AA development in mouse models and whose levels correlate well with treatment responses in patients ^E4. However, although IL-12 is the archetypal Th1 polarizing cytokine, several reports have shown that Th1 cells and a Th1 environment can be established in the absence of IL-12 or IL-12 responsiveness^E5-E7. This may indeed be the case for human and murine AA. Our data in an AA mouse model, in which disease is induced by transfer of activated T cells, and in human AA, where T cell attack of the hair follicle is ongoing, indicate that neutralizing IL-12, as well as IL-23, is ineffective at halting AA.

While in this series of 4 patients with alopecia universalis ustekinumab failed to demonstrate hair regrowth, there are reports of drug efficacy. In a series of 3 patients treated with at least 3 doses of 90 mg, there was significant change in SALT score in one patient with baseline SALT score 100% who was treated for 1 year⁸. In another series of 3 patients, there were only small changes in baseline SALT scores during treatment^E1. Meanwhile, other reports have described the development of AA coinciding with use of ustekinumab to treat other conditions^{E8, E9}.

The limitations of this study include the small number of patients and that all patients exhibited complete scalp hair loss and, therefore, may be more refractory to treatment than those with localized disease, though, as indicated, patient 1 subsequently had a complete response to tofacitinib. Moreover, we emphasize that heterogeneity may exist among alopecia areata patients, and that different patients may respond to different treatments. With regard to the preclinical experiments, although these antibodies were successful in substantially depleting their targets in the circulation, it is formally possible that these antibodies did not sufficiently modulate local cytokine levels at the interface between cells. Finally, treatment of established disease in the mouse model may be more reflective of the human scenario in which patients get treated after sign and symptoms emerge; our results indicating that IL-12/IL-23 neutralization is unable to prevent the development of murine AA, however, strongly supports that IL-12/23 neutralization is unlikely to prevent ongoing and previously established disease in this model. This is further supported by our complementary human data.

In summary, our data from an AA murine model and human patients do not support the use of ustekinumab for the treatment of AA; randomized controlled trials would lend stronger support to the efficacy or lack thereof of this treatment for AA. The C3H/HeJ mouse model of AA from which JAK inhibitors have emerged will hopefully identify additional therapeutic targets and treatments.

METHODS

Mice

Mice were kept in pathogen-free conditions at the University of Iowa Animal Facility. All animal procedures were conducted according to the NIH Guide for the Care and Use of Laboratory Animals and under University of Iowa Institutional Animal Care and Use Committee-approved protocols. C3H/HeJ mice were purchased from The Jackson Laboratory. Alopecia areata (AA) was induced by adoptive transfer of in vitro-activated cells from skin-draining lymph nodes (LN) of mice with AA^E10. Briefly, skin draining LNs were collected from mice that developed AA spontaneously. LN cells were passed through a cells trainer to generate a single cell suspension. Cells were seeded in 24 well plates with RPMI media (RPMI 1640 with 10% FBS, 2mM GlutaMAX and 100 U/mL Penicillin/Streptomycin) and cultured with CD3/CD28 Dynabeads (ThermoFisher Scientific, USA) and IL-2 (Tonbo Biosciences, USA) for 5 to 7 days. Six million cells were injected intra-dermally into the dorsal skin of twelve-week-old C3H/HeJ mice. In some cases, previously induced mice were used as donors.

Antibodies

At the time of adoptive transfer of in vitro activated LN cells, mice were treated with 250 μg of antibodies injected interperitoneally twice a week with anti-mouse IL-12 mouse p40 subunit (clone C17.8, BioXcell, USA), which has been previously shown to effectively neutralize p40 in vivo^{E11, E12}, anti-mouse IL-23 p19 subunit (G23-8, Thermo Fisher Scientific, USA)^E13, or IgG1 control (HRPN, BioXcell, USA) until the end of the experiment.

ELISA

Skin samples were collected, weighed, submerged in 1mL of RIPA buffer (Millipore Sigma, USA) and frozen at −80°C. Cells were dissociated with Gentle MACS Dissociator (Miltenyi Biotec, Germany). Skin and serum were analyzed by ELISA for IL-23 p19 (capture antibody clone 5B2 and detection antibody clone C17.8), IL1-2/IL-23p40 (capture antibody clone C15.6 and detection antibody clone C17.8), and IL-12 p75 (capture antibody clone C18.2 and detection antibody clone C17.8) according to the manufacturer’s (Thermo Fisher Scientific) instructions.

Flow Cytometry

Skin was dissociated with collagen Type 3 (Sigma-Aldrich, USA) and passed through a cell strainer. The single cell suspension was blocked with anti-mouse FcR CD16/CD32 (Tonbo Biosciences, USA) and stained for the indicated surface markers. Data collection was conducted using a LSR II cytometer (BD Bioscience, USA) and analyzed with FlowJo VX-10 software.

Statistical analysis

Statistical significance was determined using analysis of variance (ANOVA) followed by Tukey post hoc test.

Abbreviations:

AA

alopecia areata

AU

alopecia universalis

JAK

janus kinase

IL

interleukin

NKG2D

natural killer group 2D

Th

T helper

SALT

severity of alopecia tool