Trypsin shaving is a targeted proteomic method for identifying cell-surface exposed proteins on bacterial cells. For the identification of redox-active cell-surface proteins, trypsin-shaving datasets can be matched with surface-enhanced Raman spectra of intact cells to identify the cofactors associated with the cell-surface proteins. Together, these approaches could help resolve questions about the presence of cell-surface electron transport components in environmental microorganisms, especially microbes that oxidize and reduce metals and metalloids as electron donors and acceptors.
1 The importance of cell-surface redox-active proteins in environmental microbes
In the environment, phylogenetically diverse microorganisms catalyze the oxidation and reduction of metals . In some cases, these processes are carried out as part of a respiratory metabolism, but in other cases, they may be inadvertent . In the dissimilatory metal-reducing bacteria (DMRB) Geobacter, Shewanella , and Thermincola , multiheme c-type cytochromes (MHCs) have been shown to form electron conduits across the cell envelope from the quinone pool in the inner membrane to the cell surface where electrons are transferred directly to insoluble extracellular electron acceptors [3, 5-7]. In Fe(II)-oxidizing bacteria, the metal is the electron donor, and, in order to gain energy from the metabolism, the microorganism must efficiently accept electrons from the metal and avoid toxic consequences associated with metal accumulation inside the cell. Some species of bacteria, including Acidothiobacillus ferroxidans and Rhodopseudomonas palustris TIE-1, apparently oxidize Fe(II) through MHCs that are associated with the outer membrane or cell surface [2, 8], however, it remains to be determined how a variety of other species oxidize metals and whether or not cell surface redox active components are involved [2, 8].
Given their importance in electron transfer to and from metals, identification of cell-surface redox-active proteins is essential in developing a mechanistic understanding of microbial metal oxidation and reduction metabolisms. While genetic and bioinformatic tools can aid in the identification of proteins with putative extracytoplasmic localization, direct proteomic measurement of a cell surface associated protein can provide definitive identification. Furthermore, redox-active proteins frequently have cofactors with distinct chromophores, offering a spectroscopic handle for detection in biological samples. Herein, we highlight an approach that may be used to search for cell surface associated proteins with redox-active cofactors, namely, correlating trypsin-shaving proteomic data with surface-enhanced Raman (SERS) and surface-enhanced resonance Raman (SERRS) spectra of intact cells. This approach is technically straightforward, requires relatively small sample volumes and can be applied to diverse cell architectures [4, 9-13].
Correspondence concerning this and other Viewpoint articles can be accessed on the journals’ home page at:
Correspondence for posting on these pages is welcome and can also be submitted at this site.
2 Approaches to identifying cell surface proteins
Subcellular fractionation of bacterial cells with glycolytic enzymes, sugars, and differential solubilization with detergents is widely used to access “deeper” information within the proteome [14-17]. Such fractionation approaches can be used to enrich for cell wall or outer-membrane, periplasmic, inner-membrane, and cytoplasmic components  for subsequent analysis by LC-MS/MS [19, 20], and have improved our understanding of protein localization in environmentally important microbes [5, 6, 18]. Although sophisticated fractionation protocols can expand the observable proteome, fractionation has the problem of cross-contamination. For example, the presence of a protein in an outer-membrane fraction does not discriminate between the protein being surface-exposed or associated with the periplasmic face of the outer membrane. In many cases, this is an important distinction.
There are a number of approaches to targeting the surface-exposed proteome, or surfaceome, and a recent review by Cordwell highlights several of these . Chemical labeling of cell surface proteins with a biotinylation reagent followed by gel-based or gel-free proteomics has been used to quantitatively analyze cell surface proteins, however, this approach has the problem of cytoplasmic or periplasmic protein contamination . In recent years, a more straightforward approach, termed “trypsin shaving” [4, 13, 22], in which intact cells are treated with trypsin, has proved to be a powerful tool for identifying surface-exposed proteins. This approach uses inexpensive reagents, does not require chemical derivatization, and with appropriate controls can yield surfaceome datasets that are at least as reliable as those obtained by other methods.
3 Trypsin shaving as a tool for identifying cell surface proteins
When intact cells are treated with trypsin, surface-exposed proteins are more accessible to the protease and, as a result, will more readily be cleaved into peptides than membrane buried, periplasmic, or cytoplasmic proteins. The proteins for which more peptide counts are observed in a trypsin-shaved proteome are therefore more likely to be surface-exposed. An important consideration for trypsin-shaving experiments is the inclusion of an appropriate “false positive” control. Some researchers have emphasized the use of a control for proteins that are “shed” from growing bacterial cells, in particular, to eliminate proteins that come from lysed cells, are secreted, or are loosely associated with cell surfaces [13, 22]. Trypsin that is immobilized on beads (e.g. agarose or magnetic beads) may also be used to access only the most surface-exposed proteins and avoid the problem of protease leakage into the periplasm . It is an important consideration that proteases such as trypsin can access the periplasm through partially disrupted outer membranes. Therefore, the observation that a given protein is partially proteolyzed when intact cells are treated with trypsin does not, by itself, necessarily indicate that the protein is exposed on the cell surface.
In our laboratory, we have used a “lysed cell” condition to control for the presence of abundant cytoplasmic and periplasmic proteins in the trypsin-shaved proteome [4, 12] (Fig. 1). A protein for which more tandem mass spectral peptide counts are obtained in the trypsin-shaved condition relative to the number of counts obtained for the protein in the lysed cell condition is more likely to be surface-exposed. It follows that ranking proteins by trypsin:lysed ratio of peptide counts can recapitulate the organization of the proteins in the cell envelope . This was clearly demonstrated in a study of Thermincola potens. The proteins with the highest trypsin:lysed ratios in that organism were S-layer homology domain-containing proteins (which bind to cell wall polymers) and MHCs . Some researchers have suggested that trypsin shaving is more likely to be successful with Gram-positive bacteria, however, in our laboratory we have had success observing higher trypsin:lysed ratios for proteins with predicted extracytoplasmic localization in Gram-negative bacteria, based on the presence of N-terminal secretion sequences  (unpublished results). Antibody localization studies remain the “gold standard” for confirming surface localization, however, making antibodies against every candidate surface-exposed protein is time consuming and costly. This underscores the importance of having reliable, inexpensive, and targeted approaches for identifying candidate cell surface proteins.
4 Using SERS and SERRS to localize and identify cofactors associated with cell surface proteins
Alongside proteomic detection of cell surface proteins, identification of cofactors associated with those proteins on the surface of intact cells can be a valuable correlate. Flavins, quinones, hemes, and nonheme metals are chromophores that give rise to distinct Raman and IR spectra. As a specific example, heme proteins have distinct Raman peaks from porphyrin bending and stretching modes. These peaks can be resonantly enhanced with a laser excitation wavelength that falls within the heme absorbance spectrum (resonance Raman). Raman signals can be further enhanced by interaction of the Raman active molecule with gold or silver surfaces (Fig. 1) , and both SERS and SERRS have been used to visualize cell surface Raman-active species in a number of bacterial species [9, 10, 24, 25]. Enhancements can be as great as 104- to 1014-fold , but the SERS effect decays exponentially within a few nanometers from the metal surface, making it ideal for localizing a chromophore to one face or the other of a cell membrane or cell wall.
SERS has a promising role in culture-independent identification of bacteria in both direct spectroscopic assays  and in SERS-based immunoassays . Various kinds of SERS substrates may be used including electrochemically roughened gold electrodes, colloidal nanoparticles electrostatically attached to functionalized glass slides, or vapor-coated silver/gold films on glass slides [28-30]. Recent advances in the production of SERS substrates have facilitated the use of SERS for a wide range of analytical applications [23, 31-33].
In a series of pioneering studies, SERRS was demonstrated for bacterial chromatophore photosynthetic membrane preparations, but not for spheroplasts [34-36]. Carotenoids are present on the cytoplasmic face of the inner membrane, and only the chromatophore membrane preparations displayed SERRS carotenoid bands because they are “inside out” membrane vesicles. In contrast, despite the bacterial inner membrane being only a few nanometers thick, the SERRS spectrum of the spheroplasts was devoid of carotenoid peaks. This work is remarkable because it clearly demonstrates the distance sensitivity of the surface enhancement effect and the utility of the technique for identifying the membrane face on which a chromophore resides.
Another clever use of SERRS was to demonstrate surface localization of Raman-active chromophores in the DMRB Shewanella oneidensis MR-1 . Biju et al. correlated atomic force microscopy with SERRS to demonstrate the appearance of cytochrome-containing protein complexes on the surface of S. oneidensis cells starved for electron acceptor. The SERRS peaks observed in S. oneidensis correlated with nanoscale protrusions observed by atomic force microscopy, and matched the SERRS spectrum of cytochrome c. It seems plausible that the cytochromes observed in this SERRS study were MtrC and OmcA, the two outer-membrane cytochromes in S. oneidensis, but the proteins were not definitively identified in this study.
In a study with Geobacter sulfurreducens, Jarvis et al. were able to observe distinct SERS spectra from cells with intracellularly precipitated gold colloids and in the presence of extracellular colloidal silver . Surprisingly, despite the prevalence of cytochromes in G. sulfurreducens, the authors were unable to assign any of their SERS peaks to heme Raman modes.
Another promising use of surface-enhanced spectroscopy was demonstrated with spectroelectrochemical surface-enhanced infrared absorption spectroscopy of G. sulfurreducens cells grown on gold anodes in microbial fuel cells [37, 38]. In these studies, c-type cytochromes were identified in the surface-enhanced IR absorption spectra of intact Geobacter cells and potential sweeps enabled spectroelectrochemical determination of the redox potentials for the MHCs.
In our laboratory, to identify surface-associated MHCs in a Gram-positive DMRB, T. potens, we collected both SERRS spectra of intact cells in the presence of silver colloids with 413.1 nm laser excitation and SERS spectra of the intact cells on electrochemically roughened gold electrodes with 633 nm laser excitation. The observation of characteristic heme Raman bands in both sets of spectra, in combination with trypsin-shaving proteomic data indicating surface-exposed MHCs, provided evidence that there are surface-exposed MHCs  in T. potens.
5 Conclusions and challenges for the future
Correlation of SERRS/SERS results with surfaceomics data can both confirm the localization of well-characterized protein families with known cofactors and can identify new cofactors associated with surface proteins. Researchers studying metabolic pathways involving electron transfer components with chromophores should take advantage of the spectroscopic handle that redox protein cofactors provide for confirmation of proteomics-based localization studies.
As we have noted, one area of research that will continue to benefit from orthogonal measurements to determine the localization of outer-membrane proteins is the study of metal-reducing bacteria. Mechanistic models of bacterial electron transfer frequently implicate surface-exposed multiheme c-type cytochromes, and this has been clearly demonstrated to be the case for Geobacter and Shewanella species. However, very little evidence for surface-associated MHCs has been obtained for other species until recently. We identified surface-exposed cytochromes in trypsin-shaving experiments and surface-exposed hemes in SERS/SERRS experiments, thus demonstrating that surface-exposed, multiheme cytochromes are also likely involved in dissimilatory metal reduction in Gram-positive bacteria .
Trypsin shaving and SERS can also be used to identify cell-surface redox-active proteins in metal-oxidizing bacteria. Though a number of reports suggest that outer-membrane cytochromes are involved in Fe(II) oxidation, the evidence in support of their surface localization is not always strong. As Fe(II) is soluble and can diffuse through outer-membrane porins, it is important to distinguish between surface-associated electron transport complexes and electron transport complexes associated with the periplasmic face of the outer membrane. We have successfully used a trypsin-shaving approach to show the apparent absence of surface-exposed cytochromes or other electron transport proteins in microorganisms which oxidize Fe(II) or anthraquinones coupled to nitrate and perchlorate reduction , and we suggest that this approach can be applied to other metal/metalloid-oxidizing bacteria. When used together, trypsin shaving and SERS are complementary and targeted approaches for identifying surface-associated redox-active proteins.
Work on microbial interactions with iron was supported by grants to J.D.C. on “Microbial Enhanced Oil Recovery” from the UC Berkeley/BP Energy Biosciences Institute.
The authors have declared no conflict of interest.