Lupus: in search of the wolf

Geneticists have struggled for years to identify the mutant genes involved in lupus. A three-year-old patient, and a crucial discovery by Professor Carola Vinuesa’s ANU research team, will transform diagnosis and treatment of the autoimmune disorder.

At least 17,000 Australians suffer from some form of the multiple forms of the autoimmune disorder systemic lupus erythematosus (SLE), commonly known as ‘lupus’. Lupus is Latin for ‘wolf’, a linguistic relic of the mediaeval superstition that the characteristic, butterfly-shaped facial rash of cutaneous lupus was caused by a bite from a wolf.

A decade ago, Australian National University lupus researcher Carola Vinuesa embarked on a quest to identify the mutant genes that cause lupus.

A graduate of the Autonomous University of Madrid, Vinuesa is Professor of Immunology at ANU and head of the Immunology and Infectious Disease Department in The John Curtin School of Medical Research (JCSMR). She is also head of The Centre for Personalised Immunology, an NHMRC Centre of Research Excellence, which is now responsible for the lupus project.

It was a daunting assignment: the mutant genes that give rise to lupus in its multiple forms have proved as elusive and enigmatic as the rare lobo of the mountainous north of Vinuesa’s Spanish homeland — the Iberian wolf (Canis lupus signatus).

Linkage disequilibrium, the original time- and resource-intensive method for tracking down mutant genes, was never a realistic option for a complex disorder like lupus. The diversity of immune-system genes and potentially pathogenic mutations, plus the dearth of extended, multigeneration family pedigrees, made it practically impossible to use comparative genetics to identify chromosome segments or haplotypes harbouring candidate genes.

Vinuesa said the advent of powerful, highly parallel genome sequencing technologies five years ago was the game-changer for her ANU team.

Their groundbreaking study of a three-year-old Lebanese girl with a severe form of lupus, published last December in Arthritis & Rheumatology, shed new light on the nature of the genetic defects involved in lupus and signposted a path towards transformative change in diagnosing and treating the disease.

“Two years ago I read a Nature article saying the missing heritability in autoimmunity was not going to be due to rare gene variants,” Vinuesa said.

“We think that might not be entirely the case. It is likely that the genetic architecture of diseases like lupus and other autoimmune disorders will cover the entire spectrum: from one or a few rare variants to multiple common variants.

“Some cases will be oligogenic (involving a small number of genes), if not monogenic or digenic. Others may yet prove to be polygenic, but the contribution of rare variants with strong effects has probably been underestimated.”

Vinuesa said the advent several years ago of new technology for rapid exome sequencing galvanised her team’s search for lupus genes. Exome sequencers scan the genome, cherry-picking the protein-coding sequences of genes and skipping the rest — the ~98% of non-protein-coding DNA.

(The logic of exome sequencing is that, whereas mutations in gene promoters or other regions of non-protein-coding DNA may impair protein output, pathogenic mutations that distort protein structure and impair function invariably lurk in the select terrain of the exome.)

Beyond their hunt for the causative mutations in lupus, Vinuesa and her colleagues had more ambitious plans. “We wanted not just to identify rare variants that we thought were strong candidates for involvement in lupus, we wanted to prove causation,” she said.

“That really has been a major stumbling block — to convince everyone that these rare variants actually contribute to lupus.

“That hasn’t really been possible, except in cases where the gene is a very obvious candidate, like TREX1.

“For the many genes that have less well understood immunological functions, you cannot prove causation unless you make a mouse model.

“Again, that was very difficult to do until last year, when the new CRISPR-Cas9 DNA-editing tools came out.

“Previously, there was no mouse model bearing mutations found in human lupus.

“CRISPR-Cas9 technology now provides a very cheap, rapid way to introduce point mutations we find in lupus patients into a mouse embryo and create precise mouse models for the human disease.

“We’ve been doing that, and we have some very exciting discoveries. We haven’t published them yet, but several should be ready very soon.

“Without going into the details, we have found some beautiful examples of families harbouring two rare mutations in genes coding for proteins that interact with each other and thus operate in the same biochemical pathway.”

Vinuesa said the resulting deleterious effects on pathway function are similar to those seen in subjects who inherit two different, mutant alleles of the same gene — the phenomenon, known as compound heterozygosity, is common in, for example, cystic fibrosis.

“We found quite a number of these cases that result in aberrant expansion of autoantibody-secreting B cells. In each case, affected family members inherit a rare mutant allele that occurs at a frequency of less than 1% in the general population, but has a frequency of 6% in lupus patients.

“Each of the lupus patients has a second mutation in another protein that normally interacts with the first, and signals downstream of toll-like receptors.

“Our work shows that these disorders can easily be caused by rare gene variants, and it probably doesn’t take many of these rare mutant alleles to cause disease — as few as two or three.

“That’s exciting in itself, but being able to replicate those same defects in a mouse model is truly exciting.

“Not only can we use the mouse model to understand the pathogenesis of a particular disorder, we can also use it to trial therapies chosen according to known molecular defects. We can also now begin to stratify lupus patients, based on our new understanding of the molecular basis of the defects.

“In this particular case we found a B-cell abnormality that tracks with this form of lupus, based on a cytokine profile that associates with specific B cell mutations.

“That’s transformative, because we can now accurately predict that patients with this B cell disorder should respond to a particular monoclonal antibody (mAb) that may have been developed for a different therapeutic purpose.”

Vinuesa said the past two decades have seen a proliferation of mAb therapies developed for other immune disorders and cancer. Collectively, they constitute a comprehensive, off-the-shelf therapeutic arsenal for treating the various forms of lupus.

Some mAbs have already been used successfully to treat particular forms of lupus; the problem has been to match mAbs to particular forms of the disease, which usually requires multiple rounds of trial and error because there has been no information about the causative genetic lesions.

The achievements of Vinuesa and her team should drastically reduce the distance between bench and clinic. Once a mutation is discovered and its pathogenicity studied and confirmed in a mouse model, clinicians will be able to select a ready-made mAb, with high confidence that it will alleviate the chronic misery of a new subset of lupus patients. In some cases, it will save lives.

************************************************************

In 2009, a three-year-old Lebanese girl in Sydney was diagnosed with early-onset cerebral lupus.

Her case was brought to the attention of Vinuesa’s team. It would mark the beginning of a new era in lupus research.

They extracted the young girl’s genomic DNA from a saliva sample and sequenced her exome — a technique that restricts sequencing to the protein-coding regions.

JCSMR bioinformaticians then scanned her exome for candidate mutations, using algorithms developed in-house that focus on highly conserved regions of genes.

Mutations in highly conserved regions of important genes tend to be highly deleterious — even lethal.

“Many of the mutations we are looking for have high damage scores,” Vinuesa said. “They’re usually rare and highly deleterious, whereas common mutations tend to be tolerated — that’s why they’re common.

“If a mutant allele occurs at a frequency of less than 1% in the general population, we know it is less likely to have been purified by natural selection. In the case of de novo mutations, they may not be present in either parent — only in the proband (the original, affected individual).”

Vinuesa said the highest-scoring variant segregating with disease was a homozygous mutation in TREX1, a gene that had not been previously associated with paediatric lupus. However, heterozygous mutations had been shown to cause adult chilblain lupus. Homozygous mutations in TREX1 had also been shown to cause Acardi-Goutiere’s syndrome, which typically causes inflammation of the brain.

In unaffected individuals, identical TREX1 molecules pair up to form a homodimer, the protein’s active form. Vinuesa’s team determined that the young girl’s mutation, at residue 97, results in an arginine→histidine substitution that lies within in the interface region between the two molecules. It doesn’t disrupt the bond, but drastically reduces the efficiency of the dimer’s DNA-degradation activity.

The young Lebanese girl had presented with early-onset cerebral lupus; her symptoms included elevated antibody titres against double-stranded DNA fragments, indicating a defect in the cellular mechanisms that repair or degrade damaged DNA.

The girl also had haemolytic anaemia, lymphopaenia and impaired renal and liver function. She also exhibited elevated levels of interferon alpha, an inflammatory cytokine that was likely to be the primary cause of the inflammation causing her cerebral symptoms.

At age four, she had developed partial paralysis of the right side of her body. Magnetic resonance imaging of her brain revealed partial occlusion of several major arteries and widespread inflammation of medium-sized blood vessels.

There was no known history of lupus in the family, but there was an important familial clue. The girl’s parents were first cousins — an ancient tradition of marriage between first cousins is still quite common in many Middle-Eastern cultures.

In a cruel throw of the Mendelian dice, the girl had inherited the same, malfunctioning allele from both parents; they and her siblings were unaffected, indicating they had inherited at least one normal allele of the same gene, compensating for any deficiency due to the mutant allele.

Based on the ANU team’s findings, the girl’s clinicians are seeking approval to change her current treatment of six drugs with significant toxicity to an anti-Ifα monoclonal antibody, which is still in clinical trials and not yet available for individual patient use.

************************************************************

Carola Vinuesa is a plenary speaker at the Australasian Society’s of Immunology’s New Horizons conference, to be held in Canberra from 29 November to 3 December.