21-Hydroxylase-Specific CD8+ T Cells in Autoimmune Addison’s

Disease Are Restricted by HLA-A2 and HLA-C7 Molecules

By Alexander Hellesen, et al.
Presented by Louise, Makayla, Mika, Rino
Autoimmune Addison’s Disease (AAD)
● Characterized by inability to produce adequate amounts of steroid
hormones (adrenocortical insufficiency)
○ Immune-mediated destruction of the hormone-producing cells in the adrenal
cortex
○ Chronic glucocorticoid and/or mineralocorticoid deficiency may result in
adrenal crisis (nausea, confusion, fever, headache, weakness, and can be
lethal)
● The progressive destruction of adrenocortical cells in AAD is mediated by
infiltrating autoreactive T cells
Background
● 21-hydroxylase (21OH) often detectable prior to the development of
clinical disease
○ Presence of autoantibodies against 21OH acts as a biomarker of disease
○ Correlates with degree of adrenal dysfunction, but their role in degree of pathogenesis
remains unclear
● Strong genetic association between AAD and the HLA locus
○ GWAS revealed AAD risk loci in several T cell related genes (AIRE, CTLA4, and PTPN22)
● Among HLA class I genes, the B*0801 allele occurs more frequently in
patients than controls and increases their susceptibility to AAD
Previous Studies
● Demonstrated by Bratland et al. T cell responses to 21OH in patients with
AAD demonstrated in form of proliferation and IFNγ secretion by
peripheral blood mononuclear cells (PBMCs)
● Prior epitope mapping suggests presence of two immunodominant
peptides recognized by CD8+ T cell
● Responses proposed to target HLA-B8-restricted octameric sequence
EPLARLEL (position 21OH431-438), since most patients carried HLA-B*0801
allele
● Study by Dawoodji et al. reported HLA-A2-restricted epitope LLNATIAEV
(position 21OH342-350)
Limitations of previous studies:
● Previous studies only looked at detection of EPLARLEL and LLNATIAEV-
specific CD8+ T cells in only two and one patient(s), respectively.
● T cell responses to exact octameric or nonameric epitope were only shown
in study using LLNATIAEV-specific T cell line or clone that was prepared from
one individual patient.

Objective of this study:

● Characterize polyclonal CD8+ T cell responses to the proposed epitopes
LLNATIAEV and EPLARLEL in larger patient cohort with AAD.
What is the significance of mapping epitopes for
specific T cell lines?
Methods
● Peripheral blood samples from 34 patients with AAD (P1-P34), 27 healthy
patients (C1-C6, C9, C11-C-30) and 3 with other autoimmune diseases (C7,
8, 10)
○ All AAD patients had high level Autoantibodies against 21OH,
confirming autoimmune etiology
○ Informed consent + approval by ethics committee

● PBMC stimulation
○ Naive T cells were stimulated with 21OH peptides LLNATIAEV,
EPLARLEL or ARLELFVVL (10ug/ml) in 24-well plate at 1.5-3.5 * 10^6
cells/ml, once 3 days till Day 9
Dextramer and Streptamer Staining
● To isolate antigen-specific T cell
populations
○ Flexible configuration for increased
avidity

● PBMC on Day 0 and Day 13

● Stained with HLA-A2*LLNATIAEV and/or HLA-B8*EPLARLEL dextramers or
HLA-C7*ARLELFVVL streptamer
● Also for CD4, CD19, CD14 and CD8
→ Flow cytometry
Why did they use dextramer staining rather than
tetramer staining?
Hint: They were isolating T cells based on REALLY specific antigen type
IFN-gamma ELISPOT and cytokine screening
● Quantification of IFN-g-producing cells upon stimulation with 21OH
peptides
● Supernatant was collected at Day 0 and 13 for Milliplex or ELISA to screen
for other cytokines in response to stimulation

Mabtech
3. ProImmune REVEAL & Rapid Epitope Discovery System
● Outsourced analysis on physical binding characteristics of actual
peptides
○ Synthesis of peptides based on sequences
○ Kinetic analysis + flow cytometry
● Peptides covering Pep34 sequences (= type of 21OH) were
synthesized and tested for the ability to stabilize HLA-B0801

4. In Silico Predictions of HLA-C*0701-Peptide Interactions

● NetMHC 4.0 Server to predict binding of 21OH peptides to HLA class I
molecules
● pyMOL Molecular Graphics System for HLA-C
5. Protein BLAST Search
● Determining if proteins from human pathogens contain homologous
sequences to CD8+ T cell targeted-21OH-derived peptides
○ LLNATIAEV and ARLELFVVL were used - only microbial sequence homology to
both were considered
○ Blasted against a list of the most common microbial pathogens known to
colonize humans in Norway

Statistics
● Non-parametric Mann-Whitney U-test to compare frequencies of T cell
populations between patients and controls
○ Null hypothesis = two independent group have same distributions
○ Based on median (t-test is based on mean)
Result #1: High frequency of A2-specific CD8+ T cells in AAD patients
P = patient, C = control ● Frequency of A2-specific CD8+ cells
was significantly higher in Ps than in
Cs (Figure 1A, 1C)
● Even among Ps, A2-specific cells
were significantly more frequent
than B8-specific cells
● No significant difference in
frequency of B8-specific cells
between Ps and Cs (Figure 1B)
Notes: C1 and C9 had surprisingly strong
expansion of A2-specific cells but did not
test positive for 21OH autoantibodies.
P8 had more B8-specific cells but only at
frequencies 10-100 fold lower than for A2
Result #1: High frequency of A2-specific CD8+ T cells in AAD patients

*Fixed gate!

● A2*LLNATIAEV (Figure 1E)

○ 0.592% (C9) vs 13.955%
(P7) of CD8+ cells are
A2-specific
● B8*EPLARLEL (Figure 1E)
○ 0.009% (C2) vs 0.018%
(P4) of CD8+ cells are
B8-specific
 Result #2: Increased IFNγ response to LLNATIAEV (A2-restricted
epitope) in AAD patients
● IFNγ response to the A2-specific
epitope (LLNATIAEV) was greater
among AAD P’s (Figure 2A)
● On day 0, AAD P’s had significantly
higher IFNγ SFC counts than C’s
(= more IFNγ-secreting cells)
● IFNγ response to B8-specific
epitope (EPLARLEL) was not
significantly different between P’s
P = patient, C = control and C’s (Figure 2B)
Result #3: Increased IFNγ response towards Pep5 in AAD patients
(ARLELFVVL, 21OH434-442)
● Pep34 = 21OH430-447
● Pep5 = 21OH434-442
● Pep2 = 21OH431-438 (B8)

● 5/10 patients demonstrated an IFNγ

response above background to
B8
Pep34 (Figure 3)
● 4/5 patients demonstrated a clear
IFNγ response to Pep5 (ARLELFVVL)
● Low IFNγ response to B8-restricted
peptide (Pep2)!
Result #3: Increased IFNγ response towards Pep5 in AAD patients
(ARLELFVVL, 21OH434-442)

Pep34 GEPLARLELFVVLTRLLQ
Pep5 GEPLARLELFVVLTRLLQ
Pep2 (B8) GEPLARLELFVVLTRLLQ
Result #4: AAD patients have more IFNγ-producing
C7*ARLELFVVL-specific CD8+ T cells
● C7*ARLELFVVL-specific CD8+ T
cells were present in >70%
of 13 tested patients
● AAD P’s had a significantly
greater % of C7-specific CD8+
T cells than C’s (Figure 4A)
● After expansion, all P’s had
notable levels of C7-specific
cells, while all C’s had few
C7-specific cells (Figure 4B)
P = patient, C = control
Result #4: AAD patients have more IFNγ-producing
C7*ARLELFVVL-specific CD8+ T cells
● C7*ARLELFVVL (Figure 4C)
○ 0.052% (C28) vs 1.53%
(P19) of CD8+ cells are
C7-specific
● AAD patients had
significantly higher IFNγ
responses (IFNγ SFC
counts) to both
C7*ARLELFVVL and Pep34
than controls (Figure 4D)
Result #5: High sequence homology between several human pathogens
and A2- & C7-restricted 21-OH epitopes

Highest sequence homology found in

bacteria (intracellular and extracellular) → What does this suggest about how
and parasitic helminths that are known autoreactive 21OH-specific T cells are
to colonize humans. activated in AAD development?
Conclusions and Discussion
● Mechanism behind activation of autoreactive 21OH-specific T cells is unknown
○ Homology sequencing suggests molecular mimicry → environmental factor
● Identification of two dominant 21OH epitopes targeted by CD8+ T cells
○ ARLELFVVL restricted by HLA-C7
■ First HLA-C7 restricted epitope to be reported for an autoimmune disease
○ LLNATIAEV restricted by HLA-A2
○ No evidence for EPLARLEL restricted by HLA-B8
● Further support of the role of CD8+ T cells in the pathogenesis of AAD

What are some limitations and strengths of this study?

 Strengths
● Thorough discussion of reasoning that
led to specific investigations
○ Utilized and credited previous research
● Detailed methods
○ Multiple methods testing interaction
between TCR::MHC::peptides
○ Streptamer and dextramer staining for
greater specificity
● Used a strict, fixed gate to distinguish
dextramer-positive from negative cells
○ Greater confidence in their statistics
(reduced false positives)
● Did not use animal models
Limitations
● Limited by its small sample size
(34 patients)
○ Difficult to say if this is large or
small due to the rarity of AAD
● Limited by geographical location
○ Patients and control participants
were both from Norway
● Most patients have been
diagnosed with other
autoimmune diseases (e.g.
hypothyroidism, T1D, vitiligo, etc.)
could skew results
Future Studies
● Cytokine production of 21OH-specific T cells
○ Variable production of pro-inflammatory cytokines IL-6, TNFa, MIP-1a and MIP-1b in
response to LLNATIAEV → Pro-inflammatory phenotype?
○ Similar cytokine patterns found in PPI-specific CD8+ T cells in T1D patients!
○ Possibly a direct impact on the function/viability of adrenocortical cells?
● Mechanism behind autoreactive 21OH-specific T cell formation
○ Role for bacteria/pathogens in activating T cells in AAD patients
○ Is it actually molecular mimicry?
● Possibility of ARLELFVVL promoting expression of the HLA at a higher level
in adrenals than other tissues
○ Stabilize HLA-C7 cell surface expression? → mechanism for how adrenocortical cells are
targeted?
● Targeting CD8+ for AAD treatments or diagnosis
Reference
Hellesen, A., Aslaksen, S., Breivik, L., Røyrvik, E.C., Bruserud, Ø., Edvardsen, K.,
Brokstad, K.A., Wolff, A.S.B., Husebye, E.S., & Bratland, E. (2021).
21-Hydroxylase-Specific CD8+ T Cells in Autoimmune Addison’s
Disease Are Restricted by HLA-A2 and HLA-C7 Molecules. Front.
Immunol. 12:742848. doi: 10.3389/fimmu.2021.742848