The potential of 15N metabolic labeling for schizophrenia research

Abstracts

Pesquisas em psiquiatria ainda necessitam de estudos não dirigidos por hipóteses para revelar fundamentos neurobiológicos e biomarcadores moleculares para distúrbios psiquiátricos. Metodologias proteômicas disponibilizam uma série de ferramentas para esses fins. Apresentamos o princípio de rotulação metabólica utilizando 15N para proteômica quantitativa e suas aplicações em modelos animais de fenótipos psiquiátricos com um foco particular em esquizofrenia. Exploramos o potencial de rotulação metabólica por 15N em diferentes tipos de experimentos, bem como suas considerações metodológicas.

Esquizofrenia; proteômica quantitativa; rotulação metabólica de 15N; biomarcadores; G72

The potential of ¹⁵N metabolic labeling for schizophrenia research

Michaela D. Filiou

Proteomics and Biomarkers, Max Planck Institute of Psychiatry, Munich, Germany

Address correspondence to

ABSTRACT

Psychiatric research is in need of non-hypothesis driven approaches to unravel the neurobiological underpinnings and identify molecular biomarkers for psychiatric disorders. Proteomics methodologies constitute a state-of-the-art toolbox for biomarker discovery in psychiatric research. Here we present the principle of in vivo ¹⁵N metabolic labeling for quantitative proteomics experiments and applications of this method in animal models of psychiatric phenotypes, with a particular focus on schizophrenia. Additionally we explore the potential of ¹⁵N metabolic labeling in different experimental set-ups as well as methodological considerations of ¹⁵N metabolic labeling-based quantification studies.

Keywords: Schizophrenia, quantitative proteomics, ¹⁵N metabolic labeling, biomarker, G72.

The need for quantitative proteomics in schizophrenia research

In the last years, the development of holistic approaches such as genomics, transcriptomics, proteomics and metabolomics has given rise to quantitative, non-hypothesis driven research applications. In this regard, populations of genes, proteins or metabolites of two states (e.g. disease vs. control) can be compared to identify expression level differences relevant to the observed phenotypic alterations. As the proteome of an organism can reflect phenotypic changes at the molecular level, proteomics constitutes a valuable tool to investigate the underlying mechanisms involved in psychiatric disorders.

Research in schizophrenia suffers from a lack of molecular correlates for the observed behavioral alterations and disease symptoms. Molecular biomarkers for schizophrenia can aid the assessment of predisposition risk, the accurate subcategorization of patients, the monitoring of disease progression and the discovery of novel therapeutic targets. To this end, quantitative proteomics has the potential to provide sensitive molecular biomarker information and therefore offer valuable insights for schizophrenia prognosis, diagnosis and treatment.

¹⁵ N metabolic labeling

A number of quantitative proteomics approaches are available and applicable to schizophrenia research¹. Among them, in vivo ¹⁵N metabolic labeling holds great potential for studies in animal models as well as in patient cohorts. The principle of ¹⁵N metabolic labeling is based on the introduction of the stable nitrogen isotope ¹⁵N in an organism through either a ¹⁵N-labeled diet or ¹⁵N-labeled growth media. The ¹⁵N-labeled protein population (labeled specimen) is then compared to a protein sample that contains only natural-abundance isotopes (unlabeled specimen). The introduction of the ¹⁵N isotope label results in a predictable mass difference between a labeled peptide and its unlabeled counterpart. Relative protein quantification is enabled by comparing the signal intensities of the unlabeled/labeled peptide pairs of a given protein. In vivo metabolic labeling methods provide high accuracy compared to other existing quantitative proteomics approaches because the labeled and unlabeled specimens are combined prior to sample preparation, thus avoiding biased errors during experimental handling. Apart from protein quantification, ¹⁵N metabolic labeling allows the study of protein turnover in vivo through assessing ¹⁵N incorporation levels at different time points².

Applications of ¹⁵N metabolic labeling in animal models of psychiatric phenotypes and patient specimens

¹⁵N metabolic labeling has been applied to a plethora of model organisms ranging from bacteria to rodents and was recently used to label mouse models of disease³. In the context of psychiatry, a ¹⁵N metabolic labeling protocol via a bacteria-based diet was established and applied to study the mouse model of high (HAB), normal (NAB) and low (LAB) anxiety-related behavior⁴. ¹⁵N-labeled NAB mice were used as internal standards and quantitative proteomics studies in cingulate cortex, hippocampus and plasma were performed to compare HAB and LAB mice. These proteomics analyses revealed an involvement of mitochondrial and immune system-related pathways in the modulation of anxiety-related behavior^5,6.

¹⁵N metabolic labeling has also been applied to a mouse model of schizophrenia. The primate-specific G72/G30 locus is one of the most replicated findings in schizophrenia genetic studies. However, the function of the corresponding G72 protein remains to a large extend unknown. To investigate the function of the G72 protein in vivo, transgenic mice that carry the G72/G30 gene locus and express the G72 protein were generated⁷. The G72/G30 transgenic mice exhibited schizophrenia-like symptoms including increased compulsive behaviors, impaired locomotor coordination, increased sensitivity to phencyclidine, impaired odorant discrimination and learning deficits^7,8. To compare the cerebellar proteomes of G72/G30 transgenic mice and wild type controls, ¹⁵N-labeled CD1 mice were used as internal standards and several proteins related to affected molecular pathways in schizophrenia were found differentially expressed⁹.

Besides animal models, ¹⁵N metabolic labeling can be employed to examine human specimen by labeling cell lines of human origin and using them for quantitative comparisons with patient material. Human cell lines can be grown in ¹⁵N-labeled media and the deriving labeled protein populations can then be utilized as internal labeled standards for pair wise comparisons of patient and control groups. This way the in vivo metabolic labeling at the whole organism level is circumvented while the comparative proteomics analysis still benefits from the high quantification accuracy achieved by this method.

Methodological considerations

When applying ¹⁵N metabolic labeling to schizophrenia research, several considerations need to be taken into account. The cost of the ¹⁵N-labeled diet is high and long labeling periods are required for complex organisms (e.g. rodents) to achieve high ¹⁵N incorporation rates for quantitative proteomics comparisons. Importantly, the introduction of ¹⁵N may have an effect on the behavioral phenotype⁴ or on protein expression levels¹⁰. As a consequence, labeling controls (either using internal labeled standards or reciprocal labeling) should be implemented in the experimental design to avoid artifacts caused by the ¹⁵N isotope affecting relative protein quantification results. Although the cost of metabolically labeling whole organisms with ¹⁵N is not negligible, the method eventually results in ¹⁵N-labeled material (e.g. organs, blood, brain tissue etc. ) that can be used for a great number of different quantitative proteomics experiments. Given the high accuracy, fewer biological replicates compared to other less sensitive quantitative methods may be required to achieve accurate quantification results. Notably, employing ¹⁵N-labeled cell lines as internal proteome standards drastically reduces costs and labeling time to achieve high ¹⁵N incorporation, enabling a routine application of this methodology. It should be also noted that the computational challenges concerning the assessment of fully or partially ¹⁵N-labeled spectra have to a large extend been addressed by the development of appropriate software and optimized data processing workflows, which have made high-throughput data analysis possible^11,12. A detailed methodological evaluation of ¹⁵N metabolic labeling in comparison to other quantitative proteomics methods is discussed elsewhere¹³.

Conclusion

Taken together, in vivo ¹⁵N metabolic labeling provides a powerful and versatile tool for schizophrenia research that can be used both for animal model and patient-based studies. The high quantification accuracy achieved by this method may shed light on new molecular entities relevant for schizophrenia etiology and contribute to the discovery of protein biomarkers. Additionally, the study of protein turnover by ¹⁵N metabolic labeling may pinpoint protein metabolism mechanisms pertinent to the pathophysiology of schizophrenia.

Acknowledgments

The author thanks all present and past members of the Proteomics and Biomarkers Research Group at the Max Planck Institute of Psychiatry for insightful discussions and Chris W. Turck and Giuseppina Maccarrone for critical reading of the manuscript.

The author declares no conflict of interest

References

• 1. Filiou MD, Turck CW, Martins-de-Souza D. Quantitative proteomics for investigating psychiatric disorders. Proteomics Clin Appl. 2011;5:38-49.
• 2. Zhang Y, Reckow S, Webhofer C, Boehme M, Gormanns P, Egge-Jacobsen WM, et al. Proteome scale turnover analysis in live animals using stable isotope metabolic labeling. Anal Chem. 2011;83:1665-72.
• 3. Gouw JW, Krijgsveld J, Heck AJ. Quantitative proteomics by metabolic labeling of model organisms. Mol Cell Proteomics. 2010;9:11-24.
• 4. Frank E, Kessler MS, Filiou MD, Zhang Y, Maccarrone G, Reckow S, et al. Stable isotope metabolic labeling with a novel ¹⁵N-enriched bacteria diet for improved proteomic analyses of mouse models for psychopathologies. PLoS One. 2009;4:e7821.
• 5. Filiou MD, Zhang Y, Teplytska L, Reckow S, Gormanns P, Maccarrone G, et al. Proteomics and metabolomics analysis of a trait anxiety mouse model reveals divergent mitochondrial pathways. Biol Psychiatry. 2011;70:1074-82.
• 6. Zhang Y, Filiou MD, Reckow S, Webhofer C, Gormanns P, Frank E, et al. Proteomic and metabolomic profiling of a trait anxiety mouse model implicates affected pathways. Mol Cell Proteomics. 2011;10:M111.008110.
• 7. Otte DM, Bilkei-Gorzó A, Filiou MD, Turck CW, Yilmaz Ö, Holst MI, et al. Behavioral changes in G72/G30 transgenic mice. Eur Neuropsychopharmacol. 2009;19:339-48.
• 8. Otte DM, Sommersberg B, Kudin A, Guerrero C, Albayram Ö, Filiou MD, et al. N-acetyl cysteine treatment rescues cognitive deficits induced by mitochondrial dysfunction in G72/G30 transgenic mice. Neuropsychopharmacology. 2011;36:2233-43.
• 9. Filiou MD, Teplytska L, Otte DM, Zimmer A, Turck CW. Myelination and oxidative stress alterations in the cerebellum of the G72/G30 transgenic schizophrenia mouse model. J Psychiatr Res. 2012;46:1359-65.
• 10. Filiou MD, Webhofer C, Gormanns P, Zhang Y, Reckow S, Bisle B, et al. The (15) N isotope effect as a means for correlating phenotypic alterations and affected pathways in a trait anxiety mouse model. Proteomics. 2012;12:2421-7.
• 11. Zhang Y, Webhofer C, Reckow S, Filiou MD, Maccarrone G, Turck CW. A MS data search method for improved ¹⁵N-labeled protein identification. Proteomics. 2009;9:4265-70.
• 12. Haegler K, Mueller NS, Maccarrone G, Hunyadi-Gulyas E, Webhofer C, Filiou MD, et al. QuantiSpec Quantitative mass spectrometry data analysis of 15N-metabolically labeled proteins. J Proteomics. 2009;71:601-8.
• 13. Filiou MD, Martins-de-Souza D, Guest PC, Bahn S, Turck CW. To label or not to label: applications of quantitative proteomics in neuroscience research. Proteomics. 2012;12:736-47.

Publication Dates

• Publication in this collection
  14 Dec 2012
• Date of issue
  2013

History

• Received
  23 Sept 2012
• Accepted
  07 Nov 2012