Shotgun proteomics has been used extensively in recent years to survey the proteins of many organisms (16
). For example, this strategy has been valuable in surveying proteomic differences in different life cycle stages of several Plasmodium
). In the present study, we identified many expressed proteins of E. chaffeensis
by using shotgun proteomics, which involved fractionating purified total proteins of the pathogen originating from macrophages and tick cells and subjecting fractionated proteins to GeLC-MS/MS analysis. We also used GeLC-MS/MS and MALDI-TOF methods to survey individually picked total membrane proteins and immunogenic membrane proteins of E. chaffeensis
. We identified a total of 278 E. chaffeensis
proteins. The E. chaffeensis
genome was sequenced recently (17
), and the availability of complete genome sequence data greatly aided the interpretation of our proteomic data. Our study is the first and the most comprehensive proteome analysis of the human monocytic ehrlichiosis pathogen. This also is the first study confirming the expression of nearly one-fourth of all predicted genes of E. chaffeensis
, validating that they are functionally active genes, and demonstrates that classic shotgun proteomic approaches are feasible for tick-transmitted intraphagosomal bacteria.
One of the concerns with the proteome analysis, particularly for intracellular bacteria, is host protein contamination. We addressed this concern by using three different approaches. These included the use of a stringent method of purification and performing the protein identification searches using the E. chaffeensis genome data. Although this limits the possibility of identifying any contaminated host proteins, one cannot rule out the false identification of homologous host proteins as the pathogen proteins. To avoid this, we subjected the matched peptides to a database search against the nonredundant database. Our analysis found no evidence for the presence of host proteins in the pool of proteins analyzed. These measures validate that the proteome data reported in the present study represent proteins expressed by E. chaffeensis.
resides in tick and vertebrate host cells (11
), and the pathogen has two morphologically distinct forms: dense core cells and reticulate cells (4
). The dense core form is the infectious form, and the reticulate form is the replicating bacterium that resides in the phagosome of a host cell (31
). The pathogen may vary its protein expression to adapt to tick and vertebrate host environments, in support of its intracellular growth and survival, and also to transform from reticulate to dense core, infectious cells. E. chaffeensis
protein expression may also be altered in support of evading tick and vertebrate host responses. Intervention strategies will be most effective if targeted to proteins essential for different stages of pathogen growth and those expressed in a host cell-specific manner. E. chaffeensis
-expressed proteins in macrophages differed considerably from the pathogen originating from tick cells. Commonly expressed proteins were predominantly made from genes coding for normal physiological functions of a cell (e.g., those involved in protein synthesis, energy metabolism and the biosynthesis of building blocks [amino acids, nucleic acids, and lipids]). Macrophage- and tick cell-specific proteins included many hypothetical proteins, cell envelope proteins, and proteins with unknown function. Proteins identified in the unknown-function group included many novel proteins, such as ankyrin repeat proteins, GTP-binding proteins, zinc finger-like domain proteins, and metallopseudopeptide glycoprotease. The ankyrin repeat protein homologue from A. phagocytophilum
has recently been described as a secretory protein (18
) and is also considered to play an important role in modulating host response against the pathogen, possibly by interfering with host gene expression (1
). GTP-binding proteins and zinc finger proteins play important roles in regulating cell functions and gene expression (6
). Expression of these proteins in E. chaffeensis
suggests that they may also be important for cellular processes within macrophage and tick cells.
It is possible that some proteins identified as uniquely expressed in the present study may be expressed in both host cell backgrounds. This may be particularly true for a protein that is expressed at a very low level. Although host cell-specific expression of previously detected proteins, such as the p28-Omp proteins (36
), validates the relevance of data reported here, host cell-specific expression of a protein of interest must be confirmed by an independent method prior to initiating additional studies. Our study confirmed 55 hypothetical protein genes of E. chaffeensis
as truly representing functional genes. Proteins made from hypothetical protein ORFs may represent a unique group of E. chaffeensis
proteins that may serve as targets for novel drug and vaccine development.
Previous studies identified few proteins as membrane-associated proteins. In the present study, we identified several more membrane-associated proteins. Our findings also suggest the E. chaffeensis
membrane is very complex. Moreover, membrane protein structure appears to differ considerably in bacteria originating from macrophage and tick cells. Immunogenic proteins may also represent an important group of proteins involved in E. chaffeensis
interaction with tick and vertebrate hosts (1
). Although the proteins identified and listed in Table are likely immunogenic proteins of E. chaffeensis
, their immunogenicity needs to be validated independently prior to undertaking detailed immunological studies.
Recent studies suggest that A. phagocytophilum
, a rickettsial closely related to E. chaffeensis
, uses the TFSS to secrete proteins into host cell cytoplasm (18
). However, little is known about the TFSS in E. chaffeensis
). We identified here 14 putative secretory proteins expressed by E. chaffeensis
originating from macrophage and tick cells. In addition, two ankyrin repeat proteins were identified. It will be interesting to determine whether ankyrin repeat proteins are secreted into host cell cytoplasm by the TFSS, similar to the previous reports on A. phagocytophilum
). Similarly, other predicted secreted proteins may serve as effectors of the TFSS pathway or other, yet-uncharacterized secretory pathways. We presented evidence here for the expression of other secretory pathway proteins, such as the ABC transporter proteins and Sec-dependent and Sec-independent export proteins.
Our previous protein analysis showed that the p28-Omp14 is expressed in tick cell-derived E. chaffeensis
, whereas p28-Omp19 and p28-Omp20 are expressed in macrophage-derived E. chaffeensis
; we also reported several posttranslationally modified expressed forms of the p28-Omp proteins in E. chaffeensis
). We made similar observations in the present study when individually picked proteins from 1-DE-resolved membrane and immunogenic proteins were assessed by MS analysis. However, when shotgun proteomic analysis was performed, we identified peptide fragments for 18 of the 22 p28-Omp proteins in macrophage-derived E. chaffeensis
; in the tick cell-derived pathogen, expressed proteins from this locus were only from p28-Omp1 and p28-Omp14. The percentage of the identified sequence in the shotgun proteomic analysis was significantly higher for the p28-Omp19 and p28-Omp20 proteins for macrophage-derived E. chaffeensis
(see Table S1 in the supplemental material). Similarly, a greater percentage of sequence coverage was observed for the p28-Omp14 protein in the tick cell-derived E. chaffeensis
. Our identification of multiple p28-Omp proteins in the present study is similar to recent observations from membrane protein analysis reported by Ge and Rikihisa (14
). Previously, Zhang et al. (41
) reported the presence of antibodies against all 22 proteins of the p28-Omp locus in dogs infected with E. chaffeensis
. Further, RT-PCR analysis of E. chaffeensis
RNA isolated from in vitro cultures, in vivo studies in dogs experimentally infected with the pathogen, and A. americanum
ticks infected with E. chaffeensis
also demonstrated similar differential expression (5
). Although the p28-Omp19 in macrophages and p28-Omp14 in tick cells are the major expressed proteins of E. chaffeensis
), our present findings support previous reports that nearly all 22 p28-Omp proteins are expressed in the vertebrate host environment and only one to two proteins are expressed in the tick cell environment. Although expression appears to be high for one or two proteins made from the p28-Omp locus (p28-Omp19 and p28-Omp20), it is not clear why the bacterium has to express nearly all 22 proteins in vertebrate host cell environment. One possible hypothesis is that expression of the p28-Omp19 is critical for the pathogen's survival in vertebrate host environment and that the expression from other genes at low levels may provide nonessential targets for host immunity, enabling the pathogen to evade host response for its continued survival in a vertebrate host. This hypothesis, however, remains to be tested.
Popov et al. (31
) and Zhang et al. (42
) demonstrated that the outer membrane 120-kDa glycoprotein is expressed on the infectious dense core form of E. chaffeensis
but not in the reticulate form in vertebrate cells. In the present study, we detected this protein only in tick cell-derived E. chaffeensis
. We reasoned that the 120-kDa protein expression in macrophage-derived E. chaffeensis
is significantly low compared to tick cell-derived organisms. To validate this hypothesis, we performed a semiquantitative RT-PCR on RNA isolated from E. chaffeensis
recovered from different time points postinfection from macrophage and tick cell cultures (data not shown). Independent of the time postinfection, the RT-PCR data confirmed high levels of expression from the 120-kDa protein gene in tick cells and also suggested low-level expression in macrophage-derived E. chaffeensis
Only minor differences were noted in protein expression patterns of E. chaffeensis in vector and nonvector tick cell environments, suggesting that E. chaffeensis protein expression is altered in response to the tick cell environment. Vector and nonvector tick cells are very similar in inducing protein expression differences in E. chaffeensis that are distinct from the macrophage-grown pathogen.
is an emerging pathogen, and the functional genomic approach described here aided in the identification of many novel proteins. The availability of protein expression data will be valuable in initiating studies to define the functional significance of the expressed proteins to pathogen infection resulting from a tick bite. Functional genomic approaches, as described here, are crucial for validating genome data (17
) and also provide valuable information regarding protein expression. Thus, we provide here a foundation for future studies to map large-scale protein expression differences in the pathogen during growth in the vertebrate and tick hosts, particularly to determine whether similar differences in gene expression will be noted in vivo and also to determine what proteins may be contributing to pathogen evasion mechanisms.