Answers are grounded in this paper and cite their sources.
Many factors exist to make the study of proteins challenging at best. In sperm biology, these factors are coupled with the fact that a single human ejaculate can contain more than 100 million sperm cells at different stages of development and in various states of activity. In addition, these cells are transported in an active carrier and exchange extensively with both male and female reproductive tracts; ultimately, the stochastic noise means defining the function and malfunction of sperm proteins is a huge task. To make substantial progress in studying low-level, fast-changing, highly regulated, and hence important proteins, we must think outside the mass spectrometry instrument. Alternate splicing and polyadenylation, along with the use of differential promoters, enable multiple isoforms of a protein to be produced from a single gene, and posttranslational modification means even an individual isoform can present a heterologous population. Protein turnover and association with targeting subunits also contribute to making the human proteome both several orders of magnitude more complex than the genome and a transient entity, hence only definable at a fixed point in time under a specific set of conditions. Nonetheless, proteomics certainly has the potential to play a major role in characterizing the complex journey from gamete generation to fertilization, diagnosis of fertility problems, and helping to shift reproductive technologies toward safer, more cost-effective, less invasive methods (Aitken and Baker, 2007; Barratt, 2008). The high sensitivity, mass accuracy, and speed of the new generation of mass spectrometry instruments and technical developments in proteomics analysis are regularly discussed (Cañas et al, 2006; Mann and Kelleher, 2008). Also, researchers in the field now recognize that because of the sheer quantity and variability in quality and relevance of mass spectrometry data, there is a need to employ powerful bioinformatics and human expertise to avoid erroneous protein identification (Turck et al, 2007; Aebersold, 2009; Bell et al, 2009; Proteomics 2.0). This ironically echoes the calls from workers in sperm research during the last decade for the employment of universal standards and strict quality control for clinical semen and sperm analyses (Auger et al, 2000; Keel et al, 2000; Björndahl et al, 2004; Brazil et al, 2004; Keel, 2004; Björndahl and Barratt, 2005). Indeed, recently several good proteomics studies in human fertility have been reported (Oliva et al, 2009; Brewis and Gadella, 2010). It is, however, evident that several complicating factors are still hindering progress in such investigations. One major limitation is the “Achilles' heel” of the mass spectrometer itself. That is, despite the high speed and mass accuracy of the most sophisticated machines, selection of peptide ions for fragmentation is still based on abundance (typically top 6 per duty cycle), leaving a large proportion of ions unfragmented. The large proportion means that fast-changing, low-abundance proteins will rarely be identified. However, bioinformatics now offers the means to overcome this limitation. By capturing the information from the unfragmented peptide ions, mining these data, and programming the mass spectrometer to target and fragment these ions in a second or subsequent pass analysis, identifications may be obtained for a greater number of detected peptide masses. Initially this process was time consuming and laborious, but thankfully software developers and mass spectrometrists have pulled together to solve this important problem, with the label-free data mining package now an invaluable tool in the mass spectrometry laboratory. For sperm studies, this approach can be coupled with posttranslational labeling for more accurate quantitation. Also, steps such as high-level fractionation, exclusion of already identified ions, depletion of know proteins (Figure), and to a lesser extent alternative digestion and fragmentation methods can be beneficial. Enrichment methods (although these can often lead to loss) and targeted approaches such as precursor ion scanning for phosphopeptide detection can also result in more accurate quantitation (Neubauer and Mann, 1997; Williamson et al, 2006). . Steps toward improving protein identification. As mentioned and of equal importance is the problem of biologic fluctuations. Cells of the same immortal tumor cell line elicit differential responses to the application of a drug, which can in some cases lead to drastically different cell fates (Cohen et al, 2008). Embryonic stem cells, historically looked upon as identical, are now recognized as heterogenous populations (Roeder and Radtke, 2009). Not surprisingly, the heterogeneity of sperm cells within a single ejaculate creates incredible variation such that single-cell studies are imperative to gain accurate knowledge of function and malfunction in the sperm cell. Single-cell imaging and biochemical measurements of human sperm cells are already routinely carried out (Harper et al, 2005; Lefièvre et al, 2009; Smith et al, 2009). This work would be complemented by single-cell protein and small-molecule profiling to further elucidate signaling pathways associated with key mechanisms throughout the fertilization process and ultimately reveal which are failing in specific infertility-related pathologies. Part of the work of the Scottish RaSoR Consortium focuses on advancing mass spectrometry capabilities to enable single-cell proteomic profiling. Jon Cooper and his team, specialists in microfluidics and nanoengineering at the University of Glasgow, Scotland, have designed and fabricated lab-on-a-chip–based technology that uses dielectrophoresis, electroporation, and affinity capture for isolating, lysing, and detecting single cells, respectively. The technique was demonstrated in a study in which biotinylated anti–β-actin was captured on streptavidin-coated latex microspheres after lysis of human epithelial carcinoma (A431) cells, securing green fluorescent protein–tagged β-actin for observation by confocal microscopy (Sedgwick et al, 2008). Other researchers have demonstrated the use of a pulsed microbeam for cell lysis within a microfluidics system (Lai et al, 2008). This system has 2 major benefits: high-speed cell lysis that could enable observation of posttranslational modifications or temporal studies in a single cell and targeted lysis of a specific area. Because single-cell capture is routinely performed in sperm studies and effective lysis methods (laser, electrical, and chemical) are becoming available (Brown and Audet, 2008), the remaining challenge is the development of mass spectrometry methods to obtain complete protein and small-molecule profiles for individual cells. The need to use a bioinformatics approach to exploit the potential of proteomics technologies and to develop tools to enable single-cell sperm proteomics studies has been discussed. Although still in development, the necessary technology for sensitive analysis of the spatial and temporal dynamics of sperm proteins should be available in the near future. Such technologies could no doubt help increase our knowledge of the workings of the sperm cell itself, decipher the complex signaling involved in human reproductive processes, and enable the identification of indicators of the diseased state and thus selection of healthy sperm cells for assisted conception. However, what must not be forgotten is the need for rigorous screening to establish the level and dynamic range of proteins in the “healthy form” without which we cannot begin to make accurate inferences or attempt to identify biologic markers indicative of any fertility-related pathologic state. Moreover, prerequisite to all of this is true interdisciplinary collaboration encompassing researchers in reproduction, technologists, clinicians, and the expertise of the assisted conception unit.
No takes yet. Share an insight, caveat, or question.
Senga K. Oxenham (2010) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: