Our story
The story of de novo purine synthesis (DNPS) research in Prague started in 1991, when Ivan Šebesta and Jakub Krijt visited the Purine Research Laboratory at Guy’s Hospital in London, led by H. Anne Simmonds, one of the pioneers in the field of inherited disorders of purine metabolism.
They returned to Prague not only with new knowledge and experience, but also with urine samples from patients with different inherited metabolic disorders, including a patient with adenylosuccinate lyase (ADSL) deficiency. At a time when reference materials for such rare diseases were difficult to obtain, these samples provided something particularly valuable – a real biochemical fingerprint of the disease.
Back in Prague, Jakub Krijt, together with Stanislav Kmoch, began analysing urine samples from patients with unexplained neurological disorders using an HPLC system. In some samples they recognized the same unusual pattern of succinylpurines that had been present in the reference urine from the patient with ADSL deficiency.
This observation led to the diagnosis of the first Czech patients with ADSL deficiency. The first Czech cases were reported in 1994, and soon five patients had been identified. Their clinical and biochemical findings were published in 1997.
From metabolites to genes
Ivan Šebesta and Jakub Krijt played a key role in establishing the biochemical analysis of purine metabolites and the diagnosis of purine disorders in Prague. Jakub Krijt subsequently continued to develop the analytical side of ADSL research, including methods for detecting and interpreting the characteristic succinylpurines of the disease.
Stanislav Kmoch moved the research further towards molecular genetics and the functional characterization of ADSL deficiency. A major milestone came in 2000, when the Prague group published the complete human ADSL cDNA and its alternatively spliced isoform, described the structure of the human ADSL gene, identified disease-causing variants in patients and studied the functional properties of the corresponding mutant proteins.
Marie Zikánová joined this developing research line as a student in 1997. Her first work focused on the properties of normal and mutant ADSL proteins identified in the first patients. These studies raised a question that would influence much of the subsequent research: why can defects in the same enzyme lead to remarkably different clinical phenotypes?
From ADSL to the whole DNPS pathway
Answering this question gradually required looking beyond ADSL itself. The research expanded from one enzyme to the entire de novo purine synthesis pathway.
Many DNPS intermediates were not commercially available, so they had to be prepared and purified in the laboratory. Analytical methods were developed for their detection in biological material, and later replaced and extended by sensitive targeted LC–MS/MS metabolomics. This made it possible to study not only the characteristic metabolites of ADSL deficiency, but intermediates throughout the whole DNPS pathway.
At the same time, cellular models of individual DNPS defects were developed. CRISPR/Cas9 technology allowed individual enzymes of the pathway to be disrupted under defined conditions and the resulting changes in metabolite accumulation, metabolic flux and cellular physiology to be studied systematically.
The question had now changed. Instead of asking only how known DNPS disorders work, it became possible to ask whether other human diseases of this pathway were still waiting to be discovered.
The purinosome changes the view of DNPS
Another important step came from studying how DNPS enzymes are organized inside the cell. These enzymes do not necessarily function as isolated molecules dispersed throughout the cytoplasm. Under conditions of increased demand for purines, they can assemble into a dynamic multienzyme complex known as the purinosome.
In 2012, the Prague group demonstrated purinosome formation at the level of endogenous proteins in several human cell types, including primary human fibroblasts. Importantly, fibroblasts from patients with ADSL deficiency and AICAribosiduria showed impaired purinosome assembly. In ADSL deficiency, the ability to form purinosomes was related to the properties of the mutant ADSL proteins and to the clinical phenotype.
This finding changed the way DNPS disorders could be viewed. A pathogenic variant does not necessarily affect only a single enzymatic reaction; it can disturb the organization and metabolic flow of the entire pathway.
Subsequent studies using CRISPR/Cas9 cellular models showed that disruption of different steps of DNPS affects both the characteristic metabolic profile and the ability of cells to assemble a functional purinosome. The same concept would later prove important in understanding newly discovered DNPS disorders.
Searching for diseases that had not yet been discovered
The analytical methods developed over the years made systematic metabolic screening possibleand patients with unexplained neurological disorders were analysed for abnormalities in DNPS intermediates.
Most samples did not show a DNPS defect. Occasionally, however, an unusual metabolic profile appeared. Such a profile could provide the first clue that a previously unknown block existed somewhere in the pathway.
One of these stories led to PAICS deficiency.
In collaboration with international partners, patients with a defect of the bifunctional PAICS enzyme were identified and studied at the genetic, biochemical and cellular levels. In 2019, PAICS deficiency was reported for the first time as a human inherited disorder of de novo purine synthesis.
The first patients had multiple congenital abnormalities and died shortly after birth. Functional studies showed markedly reduced PAICS activity and, importantly, impaired purinosome formation in patient fibroblasts. Introduction of the normal PAICS protein restored purinosome assembly, directly linking the genetic defect with disruption of the cellular organization of DNPS.
Further patients later showed that the clinical spectrum of PAICS deficiency is considerably broader. The disease can also present with a later-onset neurodevelopmental phenotype, including epileptic encephalopathy and progressive cerebral atrophy.
And then it happened again
Several years later, targeted metabolic screening revealed another unusual profile. Patients with unexplained neurological symptoms showed increased urinary levels of formylglycinamide riboside (FGAr). The position of this metabolite in the pathway pointed towards another DNPS enzyme – phosphoribosylformylglycinamidine synthase (PFAS).
Genetic analysis identified biallelic variants in the PFAS gene, and biochemical and functional experiments confirmed their pathogenic effect. Patient fibroblasts showed reduced PFAS protein and activity and impaired purinosome formation. Expression of normal PFAS restored purinosome assembly.
In 2025, the group reported the first patients with PFAS deficiency, establishing another previously unknown inherited disorder of de novo purine synthesis.
The story continues
More than three decades separate the old HPLC analyses of the first Czech ADSL patients from today's metabolomic, genomic and cellular approaches.
What began with ADSL deficiency gradually expanded to the whole DNPS pathway. Along the way, the research contributed to understanding the molecular basis of ADSL deficiency, showed the purinosome, and led to the discovery of two previously unknown human metabolic diseases – PAICS deficiency and PFAS deficiency.
Today, the same research line continues from diagnosis and disease discovery towards a deeper understanding of metabolic regulation, cellular organization and disease mechanisms, and ultimately towards the development of targeted therapeutic approaches for DNPS disorders.