Memory bias to Flu virus globular head explains rational for universal stem[stalk]-only Flu immunogen design

Few days ago Science Translational Medicine published interesting study explaining hurdles for creating universal Flu vaccine. Universal Flu vaccine refers to a "hypothetical" Flu vaccine that can induce cross-strain neutralizing antibody response. 

There are dozen of Flu viruses that differ from each other in H and N antigen composition. However, multiple studies that followed 2009 H1N1 Flu pandemic revealed that the most Flu viruses share conservative region called stem [or stalk] region located below their globular head [main target of anti-Flu antibodies]. Antibody response to Flu stem/stalk region is not common but when induced it has potential to cross-neutralize almost every human-specific Flu viruses. Such vaccine, obviously, has a huge medical application and carries huge marketing potential too.

So why are these anti-Flu stem/stalk antibodies so rare? First clue came from serial analysis of serum samples harvested from individuals immunized second time with H1N1 vaccine [that included both globular head ad stem/stalk regions]. Unlike primary immunization, secondary vaccination re-directed immune response towards Flu globular head at the expense of flu stem/stalk region. Basically, anti-globular head antibody memory response could easily out-compete anti-Flu stem/stalk region antibody memory response [thus reducing cross-neutralizing potency].

Second and more puzzling clue came from observation that unlike anti-Flu globular head antibodies, anti-Flu stem/stalk antibodies displayed excessive non-specific binding [in an in vitro assay] to irrelevant antigens. This raises basic question whether human immune system had been "intentionally" keeping away from Flu stem/stalk response to avoid any potential cross-reactivity to self antigens. However, so far there is no evidence to suggest that this in vitro "sticky" character of anti-Flu stem/stalk antibodies plays any clinically relevant role in vivo.  

In summary, this study shows that conventional Flu vaccine design that incorporates both Flu virus globular head as well as its stem/stalk region wouldn't work to induce long-lasting cross-neutralizing universal antibodies. However, Flu stem/stalk only immunogens could work as reported earlier in two independent studies. 

Hemagglutinin-stem nanoparticles generate heterosubtypic influenza protection.

A stable trimeric influenza hemagglutinin stem as a broadly protective immunogen.

David Usharauli

ADAMTS-like protein 5 (ADAMTSL5) is a melanocyte-derived auto-antigen in psoriasis

Psoriasis is a chronic skin inflammatory disorder characterized by thickening of skin epidermal layer. Psoriasis is considered autoimmune diseases due to presence of activated T cells within epidermal layer in affected skin and clinical benefits of antibody therapy directed against cytokines such as IL-23 and IL-17. However, the exact (auto)antigen[s] targeted by T cells are not well known.

This short paper from Journal of Experimental Medicine suggests that a disintegrin and metalloprotease with thrombospondin motifs-like-5 protein (ADAMTSL5) expressed by melanocytes is a psoriasis-specific auto-antigen.

This study is continuation of earlier studies showing that (a) HLA-C*06:02 allele occurs in more than 60% of psoriatic patients, and (b) psoriatic lesion contains high frequency of Vα3S1⁺/Vβ13S1⁺ CD8 T cells. Now, using in vitro HLA transfection experiments the authors showed that Vα3S1⁺/Vβ13S1⁺ T hybridoma recognized HLA-C*06:02⁺melanocytes [but not keratinocytes or fibroblasts] in a HLA I-specific manner.

Next, using bioinformatics approach in combination with in vitro hybridoma activation assay (NF-κB-GFP readout), the authors identified ADAMTS-like protein 5 (ADAMTSL5) as a potential auto-antigen in psoriasis.

ADAMTSL5 was highyl expressed by both primary and cell line melanocytes (WM278).

Finally, the authors showed that ADAMTSL5 derived peptide stimulated psoriatic T cells but not T cells from healthy individuals.

In summary, this study showed that melanocyte-derived ADAMTSL5 represents auto-antigen in psoriasis. The authors claim that this is a first definitive report of showing psoriasis as a truly autoimmune diseases.

David Usharauli

Chromogranin A is a disease-driving auto-antigen in type I diabetes

Type I diabetes is autoimmune disease leading to destruction of insulin-secreting endocrine β islets in pancreas and glucose intolerance. NOD mouse strain represents well established model of human type I diabetes. So far, researchers identified several potential auto-antigens targeted by self-reactive T cells, out of which so far only mutation in insulin B chain (B:9–23) showed effect on disease progression.

Now, this Cutting Edge study in Journal Immunology showed that absence of another auto-antigen, Chromogranin A (ChgA), fully protected NOD mice from diabetes development.

First, the authors showed that β islets of NOD.ChgA^-/- could not activate ChgA-specific T cell clones in an in vitro assay (reactivity to other auto-antigens insulin and IAPP was intact).

Importantly, the authors showed that NOD.ChgA^-/- mice were completely protected from development of diabetes [measured as excess of glucose in urine samples].

Finally, the authors showed that NOD.ChgA^-/- mice showed minimal β islets inflammation (insulitis) but developed comparable salivary gland inflammation (sialitis), implying that Chromogranin A effect was strictly β islets selective.

In summary, this study showed that at least Chromogranin A is necessary and sufficient for development of diabetes in NOD mice. Still, it is puzzling that NOD.ChgA^-/- mice is protected from diabetes when they still expressed another dominant auto-antigen insulin [because earlier study showed that NOD ins1^-/- ins2^-/- mice were protected from diabetes development too]. If both auto-antigens are "priming epitopes" then both NOD.ChgA^-/- and NOD ins1^-/- ins2^-/- mice should have developed diabetes. Strange.

David Usharauli