Surprisingly variable “dangers, toils, and snares” faced by humans and mice

In this issue of Blood, back-to-back papers by Sepulveda et al¹  and Kögl et al²  investigate mechanisms behind the range of immunopathologies associated with syntaxin-11 deficiency and come to similar conclusions.

Cytotoxic lymphocytes (CLs) engage the secretory granule death pathway to eliminate virus-infected or transformed host cells. Cytotoxic secretory granules encapsulate perforin that, when released into the immune synapse, forms pores in the target cell membrane. These allow entry for co-secreted, proapoptotic granzyme serine proteases that then process highly specific cytoplasmic substrates to initiate an irreversible apoptotic cascade. But apart from its anti-infective and anti-neoplastic roles, this process is also critically responsible for overall immune homoeostasis: a congenital failure to deliver functional perforin is the cause of the often fatal autosomal recessive disorder, Familial Hemophagocytic Lymphohistiocytosis (FHL).³ 

It took 50 years to crack the molecular basis of FHL since its initial description: a failure of the secretory granule pathway with disordered perforin biosynthesis, trafficking, secretion, or function. The critical importance of perforin and some granzymes in protecting against pathogens in vivo and preventing postinfective hemophagocytic syndrome(s) in mice was demonstrated in the early 1990s⁴ ; a few years later mutation of the perforin gene became the first established cause of human FHL,³  accounting for 30% to 60% of all known cases. More recently, most of the remaining cases of FHL were linked to mutations in 3 other genes, which regulate cytotoxic granule exocytosis: UNC13D (MUNC13-4),⁵ STX11 (SYNTAXIN 11),⁶  and STXBP2 (MUNC18-2)^7,8 ; a link was also made to Griscelli syndrome, which results from defects of RAB27a (RAB27A).⁹  Apart from their unquestionable clinical significance, these studies have also enlightened our understanding of CL cellular and molecular biology.

Although FHL was previously considered exclusively a disease of early childhood, it is now accepted that it can also present considerably (sometimes much) later, and with a remarkable spectrum of symptoms that make definitive diagnosis difficult. The complete loss of functional perforin protein because of bi-allelic mutation invariably leads to florid FHL in infancy, but the residual activity of some missense mutations can delay onset of FHL until adolescence or later, or may produce alternative presentations such as persistent EBV-related pathology or, perhaps more commonly, a range of hematologic cancers.¹⁰  The current studies both used mouse models to reveal an important difference between PRF1- and STX11-associated FHL. The onset of FHL in patients with STX11 mutations is typically delayed by some months compared with PRF1-mutation. Sepulveda et al compared the disease-free periods for patients with bi-allelic nonsense or frame-shift mutations that result in a complete loss of function for both proteins.¹  The results were surprising: the disease-free period for STX11-null patients is considerably longer than when PRF is totally absent, or in Griscelli syndrome (mutation in RAB27a).

To investigate the progression of disease and its underlying immune basis ab initio, both groups generated stx11^−/− mice and compared them with prf1^−/− mice.^1,2  As predicted, the CL of stx11^−/− mice cannot undergo degranulation (unlike their prf1^−/− counterparts), and their targets survive. As in humans, following a significant external immune stimulus (virus infection), mouse FHL is driven by hyperactive CD8⁺ T cells, with consequent elevation of IFNγ secretion and macrophage activation. However, challenging stx11^−/− mice with lymphocytic choriomeningitis virus (LCMV) led to a major surprise. Whereas prf1^−/− succumbed to florid disease, stx11^−/− mice that generated similar viral titers survived the challenge without clearing virus. This suggests that besides regulating CL degranulation, syntaxin-11 has other roles in the immune system, which compensate progressive fatal FHL. This relatively benign outcome also suggests that syntaxin-11 deficiency in humans may be more common than currently appreciated, as milder cases may be self-limiting as a result of compensatory mechanism elsewhere within the immune system.

A comparison of rab27a^−/− and stx11^−/− mice by Sepulveda et al reveals that despite inflicting similar defects in degranulation and cytotoxic function, rab27a^−/− mice had a more severe disease phenotype.¹  Further experiments supported a previously reported role of Rab27a in antigen presentation, and were consistent with pronounced proliferation of antigen-specific CD8⁺ cells, leading to a more severe disease. It became clear that a simple deficiency of CL cytotoxiciy is not the only determinant of FHL pathology, and the role of syntaxin-11 in myeloid or other immune cell types can potentially influence disease progression.

That the expansion and up-regulation of CD8⁺ T-cell activity in stx11^−/− mice has nonlethal consequences points to fundamental differences between this form of disease and prf1 deficiency. In an unexpected twist (and unlike in prf1^−/− mice), Kögl et al show the re-stimulation of LCMV-infected animals with class I–presented viral peptides causes a significant decrease in IFNγ⁺ CD8⁺ T-cell numbers, suggesting the eventual onset of T-cell exhaustion, suppression, or apoptosis.²  In the longer term, this reduction of IFNγ⁺ CD8⁺ T-cell activity ameliorated the disease phenotype. The mechanism was confirmed by demonstrating the raised expression of several inhibitory receptors on antigen-specific CD8⁺ T cells of LCMV-infected stx11^−/− mice. Conversely, blocking inhibitory receptors PD-L1 and LAG-3 aggravated the disease in stx11^−/− and resulted in death, confirming that T-cell exhaustion was a key feature of the milder form of haemophagocytic lymphohistiocytosis. Whether the findings apply to other forms of FHL, most notably MUNC18-2 (STXBP2) deficiency, which leads to the loss of syntaxin-11 expression,^7,8  and whether these results will be recapitulated in humans remains to be seen.

A further surprise was that unlike in FHL patients, the cytotoxic activity of stx11^−/− lymphocytes could not be compensated by IL-2 stimulation. The reason is unclear, but relates to the status of the immune system (eg, largely naive in mice housed in controlled conditions but constitutively primed by antigen in humans). Irrespective of this point, these 2 fascinating studies^1,2  open up many opportunities to address further questions on the basic cellular biology of the secretory apparatus of immune cells, understanding the cross-talk between antigen presenting cells and immune effectors cells, better understanding the impact that immunoregulation has on infective and neoplastic diseases, and ultimately, the role of environmental factors in shaping the impact of apparently similar immune deficiency states. They reveal (and further confirm) the remarkable heterogeneity of FHL, and give hope that this group of diseases may be better understood, paving the way for easier diagnosis and more effective management.