Cocktail of labelled VOC-based probes for biomedicine – Volatolomix

Within the framework of this project, our targets included N-acetyl-β-glucosaminidase, β-glucuronidase, α-L-fucosidase, β-galactosidase, α-mannosidase, β-glucosidase, and sialidase. Heterosidic probes were developed, consisting of a glycone moiety, substrate for each of these enzymes, linked to ethanol (a highly volatile VOC that readily transitions to the gas phase). To differentiate the ethanol released after hydrolysis of each probe, ethanol was labeled with ¹³C and/or ²D at one or more positions. A method for preconcentrating VOCs using SPME coupled with targeted GC-MS/MS analysis (MRM mode) was developed to preconcentrate and detect all ethanol isotopes. This method enabled clear differentiation between all isotopes based on their mass fragmentation.

A comprehensive experimental study was then conducted to evaluate the affinity, sensitivity, and cross-reactivity of the probes. The detection limits of the enzymes were subsequently determined in solution.

We then developed a multiplex strategy, combining proteomics and volatolomics, to detect exoglycosidases directly in biological samples, including plasma and tissues. The objective was twofold: (a) to implement a simple, robust, and easily reproducible procedure based on adding VOC-based probes to biological samples to map enzymatic activities; and (b) to develop a proteomic protocol for extracting and analyzing enzymes in these environments. This multimodal strategy, combining proteomics and volatolomics, was designed to map hyperactive enzymes that could serve as biological targets for the development of new drugs.

Once these protocols were established, a clinical study was conducted on plasma samples (in collaboration with clinicians from the INSERM 1402 CIC at Poitiers University Hospital and the PHP Paris). This clinical research study, involving several dozen patients, aimed to assess the effectiveness of the VOC-based probe cocktail in distinguishing healthy patients from those suffering from infection or inflammation, as well as to perform longitudinal monitoring of these patients. The probe responses were correlated with the clinical and biological data of the patients.

After defining the optimal conditions for detecting volatile tracers and activating VOC-based probes, we developed an initial cocktail consisting of 4 VOC probes to identify new glycosidic biomarkers associated with tumor development. Using these VOC-based probes, we identified a target enzyme, N-acetyl-glucosaminidase, present both in vivo and in vitro in the environment of cervical and triple-negative breast tumors. We subsequently developed a prodrug activatable by this enzyme, and after in vivo trials, we successfully cured 4 out of 6 mice implanted with human mammary tumor xenografts, with no major side effects. Currently, there is no specific treatment for this type of cancer, and conventional chemotherapies do not prevent relapse. Thus, thanks to VOC-based probes, we were able to identify new therapeutic targets, leading to the development of a novel chemotherapeutic agent with unprecedented efficacy.

In parallel, we optimized a protocol for analyzing glycosidic activities in solid biopsies to assess the relevance of the probe cocktail in identifying pre-neoplastic lesions. This study was conducted using a gastric carcinogenesis model linked to Helicobacter pylori infection. We identified two enzymes whose activity correlated with the onset of early signs of tumor development. These results demonstrate the potential of VOC-based probes for the very early detection of cancerous tumors.

Finally, we developed a protocol to study the evolution of glycosidase activity during infection. We focused on SARS-CoV-2 infection and identified three glycosidases, α-mannosidase, β-glucuronidase, and N-acetyl-glucosaminidase, that were overactivated 8 days after the onset of symptoms in patients with severe COVID-19 compared to those with mild forms. These three enzymes could potentially predict disease progression. Additionally, one of these enzymes, N-acetyl-glucosaminidase, was identified as an infection marker. The other two enzymes, α-mannosidase and β-glucuronidase, were positively activated 16 days after symptom onset. Through in vitro infection assays, we observed that β-glucuronidase was not produced by infected cells, as we did not detect its activity in these cells. This enzyme is likely produced by other cells, such as immune cells. This finding demonstrates our ability to predict the progression of an acute infection toward chronic inflammation.

This project has naturally led to a large-scale clinical study involving over 220 individuals. The aim will be to demonstrate the relevance of this new approach for detecting breast, lung, and pancreatic cancers. A second phase of the study will focus on demonstrating the effectiveness of this approach in monitoring the success of therapeutic protocols for diagnosed patients. This study is set to begin in mid-2026 and will last approximately 1.5 years.

Additionally, new VOC-based probes are currently under development to detect sepsis in patients. These probes are also expected to enable rapid identification of bacterial strains and their potential antibiotic resistance.

Finally, the strategy proposed here was the subject of an invention disclosure filed in December 2020, followed by an effective patent application in 2022, with promising results. This patent has been licensed by a national start-up specializing in the development of new diagnostic tools for diseases.

With the emergence of volatolomics, i.e., the discipline that investigates VOCs produced by living systems, the human volatolome (i.e. matrix of VOCs released by an individual) has been emphasized since it can provide valuable insight into biological processes in real-time. Hence, as evidenced by canine olfaction-based studies, volatile molecules can serve as chemical tools for assessing human health status in real time in a noninvasive way. In comparison with markers based on macromolecules like nucleic acids or proteins, endogenous VOCs are safer and can be sampled noninvasively (at least for breath, sweat, sebum or urine), quickly and as often as necessary, thus facilitating their transfer to the clinic. No endogenous volatile compound has proven to be of sufficient diagnostic value so far. This can be explained in terms of both inter-individual variability, individual daily variation, and lack of standardised protocols for sample collection, preparation and analysis.
In this context, we hypothesise that a mix of exogenous VOCs that do not result from endogenous metabolic pathways but rather from metabolic-specific enzymatic conversion of a probe would overcome these issues. Thus, the objective of this project is to propose an emergent toolbox that relies on a cocktail of labelled VOC-based probes developed to diagnose and monitor pathophysiological events within a body. In this pilot project, our cocktail will target glycosidases that play key roles in many biological events, such as inflammation. Once metabolized, the probes release exogenous VOCs that will be used as chemical tracers of the biochemical activities related to inflammation. With these probes in hand, we envisage to screen enzymatic biomarkers at the molecular, cellular, tissular and organismal scale. Thereafter, these markers could be targeted by new therapeutic agents working in an integrative manner to stop disease progression.